sustainability-16-11147

Article
Enhancing Organizational Resilience: Sustainable Development

Scenarios Incorporating Disaster Impacts and AI Tools
Augustinas Maceika 1,† , Andrej Bugajev 2,*,† and Olga Regina Šostak 2,†
1 The Faculty of Mechanics, Vilnius Gediminas Technical University, Plytinės Street 25,
LT-10105 Vilnius, Lithuania; augustinas.maceika@vilniustech.lt
2 The Faculty of Fundamental Sciences, Vilnius Gediminas Technical University, Sauletekio Ave. 11,
LT-10223 Vilnius, Lithuania; olga-regina.sostak@vilniustech.lt
* Correspondence: andrej.bugajev@vilniustech.lt
† These authors contributed equally to this work.
Abstract: The intensification of human activities and the escalating impact of climate change have
increased the probability of disasters, making it important to develop sustainable development
scenarios that consider potential disaster consequences. However, disasters are indirectly represented
in the 17 Sustainable Development Goals (SDGs) and often overshadowed by other topics. This study
focuses on disaster effects in the context of sustainable development. We conducted a sociological
survey with 30 respondents from Lithuanian companies, half of which were large manufacturing
enterprises, and found that only 37% had encountered sustainable development and disaster man-
agement issues, with a similar proportion actively developing related scenarios. Although external
stakeholders often participate, 57% of the respondents indicated that their company acts indepen-
dently on these matters. Large companies rated their disaster preparedness higher (3.5/5) than SMEs
(2.9/5) or micro-companies (2.8/5). Rapid response and liquidation of disaster consequences was
deemed more important (4.5/5) than risk management and resilience-building scenarios (3.8/5).
Using correlation and mutual information analyses, we uncovered linear and non-linear dependen-
cies, showing that psychological stress among employees negatively correlates with the importance
assigned to SDGs during disasters. Additionally, we demonstrated how generative AI tools, such
as ChatGPT, can transform survey results into tailored scenarios. These findings provide practical
Citation: Maceika, A.; Bugajev, A.; insights and methodologies for enhancing organizational resilience and sustainability, even with
Šostak, O.R. Enhancing limited resources.
Organizational Resilience: Sustainable
Development Scenarios Incorporating Keywords: sustainable development; disaster management; scenario generation; artificial
Disaster Impacts and AI Tools. intelligence; ChatGPT; prompt engineering; sustainable development goals; disaster preparedness;
Sustainability 2025, 16, 11147. mutual information analysis; Lithuanian manufacturing companies
https://doi.org/10.3390/su162411147
Academic Editor: Sotirios Argyroudis
Received: 23 November 2024 1. Introduction

Revised: 12 December 2024
In recent years, the world has faced an unprecedented combination of global crises:
Accepted: 17 December 2024
Published: 19 December 2024
intensifying climate change, pandemic risks, geopolitical conflicts, and economic stagna-
tion. This makes the need for resilient and sustainable development more urgent than
ever. These overlapping challenges, often described as a “polycrisis”, have especially
severe implications for developing countries, where they hinder progress towards poverty
Copyright: © 2024 by the authors. reduction and sustainable development and exacerbate existing vulnerabilities [1].
Licensee MDPI, Basel, Switzerland. The United Nations’ 17 Sustainable Development Goals (SDGs) provide a comprehen-
This article is an open access article sive framework for addressing the global challenges [2]. However, within this framework,
distributed under the terms and the risks posed by disasters—both natural and man-made—are often addressed indirectly
conditions of the Creative Commons and may be overshadowed by other critical issues such as economic growth, health, and
Attribution (CC BY) license (https://
education. Some research results showed dramatic escalations in both natural and techno-
creativecommons.org/licenses/by/
logical disasters over the decades [3]. Recent events have demonstrated that disasters can
4.0/).
Sustainability 2025, 16, 11147. https://doi.org/10.3390/su162411147 https://www.mdpi.com/journal/sustainability

Sustainability 2025, 16, 11147 2 of 39
severely disrupt progress towards these goals, highlighting the need to integrate disaster
risk management directly into sustainable development planning [4,5].
While general instruments that increase catastrophe reactions are unquestionably
important [6], scenario analysis plays a crucial role in planning and managing the elimina-
tion of the consequences of disasters and emergencies [4]. However, developing sustainable
development scenarios that account for disaster impacts remains a labor-intensive and
resource-demanding process. This challenge is particularly acute for organizations in
developing countries, which often lack the necessary resources to conduct thorough sce-
nario analyses and incorporate long-term perspectives into disaster risk management [7].
Additionally, the complexities of long-term changes and multiple actors make disaster risk
reduction inherently challenging, requiring the incorporation of uncertainties and dynamic
factors into risk assessments [5]. As a result, there is a significant gap in the ability of these
organizations to prepare for and respond to crises effectively.
This research aims to address these challenges by identifying and analyzing the
critical factors that influence sustainable development scenarios in the context of disaster
risk. Through a combination of a literature review and sociological surveys conducted
among organizational employees, this study explores the key drivers of disaster resilience
and their interactions. Furthermore, the study investigates the potential of leveraging
modern artificial intelligence tools, such as large language models and conversational
agents, to streamline the scenario development process. By demonstrating how AI can be
used to enhance scenario planning, this research offers innovative solutions for resource-
constrained organizations, as pursued in other research [8], where conversational AI in the
form of a chatbot was utilized for disaster risk reduction.
The remainder of this paper is structured as follows. In Section 2, we provide the
background of the research. Section 3 describes the methodology of the research. Section 4
presents the results obtained from the survey and application of the analysis methodology.
Finally, Sections 5 and 6 are dedicated to the discussion and conclusion. Additionally, we
provide Appendices A–C with ChatGPT prompts and a full correlation coefficient heatmap.
2. Background of the Research

In an era characterized by escalating uncertainties due to climate change, pandemics,
and geopolitical tensions, scenario planning has become an indispensable tool for orga-
nizations aiming to navigate complex future potentials and enhance resilience. Scenario
planning involves creating plausible and detailed narratives regarding how the future might
unfold, enabling organizations to anticipate potential challenges and opportunities [9].
2.1. Scenario Planning in Organizational Strategy

The use of scenarios as an effective tool for adapting the strategic beliefs of the
managers of an organization provides a new perspective in organizational cognition.
Vecchiato [9] highlights that, although scenarios have been extensively praised, there is
still limited understanding of their effects on managers’ mental models (representations of
reality) and their impact on strategic investment decisions under uncertainty. This gap sug-
gests a need for further research into how scenario planning influences the decision-making
processes within organizations.
One of the most notable examples of scenario planning in a corporate context is Shell’s
extensive use of scenarios to guide strategic decisions. After an oil crisis, Shell adopted
scenarios widely throughout the company, evolving both in scope and process. The initial
focus on key variables such as energy demand and oil prices expanded to include economic,
political, social, and environmental factors [9]. Notably, Shell’s scenario-based planning
approach was a major success, which was recognized by many experts. Thanks to its
scenarios, Shell was better prepared for the crisis, and the company recovered more quickly
than competitors from such crises as oil embargoes and the growing interventionism of the
Russian administration [9]. Shell’s scenario planning process involved several steps:
1. Listing predetermined elements and trends in the global energy market.
2. Identifying drivers of change in the external environment that could affect the com-
pany’s competitive position.
3. Exploring likely patterns of evolution for these drivers of change.
4. Building the scenarios’ plots and narratives.
5. Analyzing the industry and market structure that would prevail in each scenario.
6. Identifying emerging sources of competitive advantage.
7. Generating and evaluating options for strategic action.
Shell developed scenarios at three levels:
• Global Scenarios: Exploring forces in the global macro-environment, including politics,
economy, society, ecology, technology, and demographics.
• Focused Scenarios: Addressing each business sector and geographic area based on the
global scenarios.
• Project Scenarios: Investigating specific investment projects with detailed information
on competitors, profitability, and risks.
These scenarios enabled Shell to anticipate changes and adapt its strategies accord-
ingly. For instance, in the early 1990s, Shell’s global scenarios focused on drivers such
as globalization and liberalization, outlining alternative futures to guide the strategic
decisions [9].
However, critiques have emerged regarding the effectiveness of Shell’s scenario plan-
ning. Zalik [10] argues that Shell’s scenarios often reflect the company’s preferred outcomes,
potentially overlooking less favorable but plausible futures. In “Scenarios to 2025”, devel-
oped by Shell, three futures were presented—“Low-Trust Globalization”, “Open Doors”,
and “Flags”. From these scenarios, the “Open Doors” scenario was clearly the preferred
option, offering stronger economic growth over time [10].
Jefferson [11] further critiques Shell’s scenario planning, highlighting issues such as
over-reliance on a few individuals without ensuring adequate succession, logical inconsis-
tencies, and failures to anticipate critical events like the oil price collapse in the mid-1980s.
He points out that considering a scenario as “unsustainable and unsuitable” when it ac-
curately describes future events is a logical absurdity. Moreover, Shell’s abandonment of
medium-term scenarios compounded the strategic errors in planning.
In response to claims that scenario planning is a “black box”, Jefferson [12] defends
the transparency of Shell’s scenario planning process. He argues that the inner workings
of scenario development are fully revealed—a “clear box evaluation”—and that evidence
of openness is available for those willing to engage with it. He further emphasizes that
developing and applying alternative scenarios can help to highlight opportunities and
challenges, assess realism, and explore compatibility with the Sustainable Development
Goals [13].
These discussions underscore the complexities and challenges inherent in scenario
planning within organizations. They highlight the need for robust methodologies that effec-
tively capture uncertainties and inform strategic decisions without bias toward
preferred outcomes.
2.2. Scenario Planning for Sustainable Development and Disaster Risk Reduction
Scenario planning has also been applied extensively in the context of sustainable
development and disaster risk reduction. The Sustainable Development Goals (SDGs)
set by the United Nations provide a comprehensive framework for global development,
but integrating disaster risk management into these goals remains a challenge. Even
the existing standards for smart community infrastructure cannot meet the new reality
requirements [14]; thus, there is a need to adapt them to meet the SDGs.
Scenario planning has been studied by authors who have focused on selected strategic
research directions, applying various research methods. One such work is by Ameli et al. [15],
which examines how the COVID-19 crisis has significantly impacted the implementation
of the 2030 Agenda for Sustainable Development in Iran. Using a fuzzy cognitive map
to specify cause–effect links among interdependent SDGs, they analyze the potential
effects of the pandemic on SDG achievement. They develop scenarios corresponding to
five proposed strategies and test them under different pandemic activation levels. Their
findings indicate that applying certain strategies at high activation levels can mitigate
the impact of COVID-19 on the SDGs, highlighting the importance of integrating disaster
considerations into sustainable development planning.
Similarly, Cernev [16] explore the effects of crossing planetary boundaries and global
catastrophic risk (GCR) events on disaster risk reduction efforts and international devel-
opment targets. Through scenario analysis, they develop distinct futures for humanity
and Earth under different levels of GCR and planetary boundary transgressions. They
identify scenarios such as “Earth Under Uncertainty”, “Global Collapse”, “Stable Earth”,
and “Earth Under Threat”, emphasizing that, without direct action, the world may be on
a pathway toward the “Global Collapse” scenario. Their work underscores the need to
integrate planetary boundaries and GCR considerations into international development
goals and disaster risk reduction frameworks.
Examples of scenarios that study the environment of a company in the context of urban
development are relevant to our topic. For example, Fang et al. [17] address the complex
interactions between urbanization and the eco-environment in China’s Beijing–Tianjin–
Hebei urban agglomeration. They develop the Urbanization and Eco-environment Coupler
(UEC), a system dynamics model that simulates the non-linear relationships between
various elements. By creating comprehensive scenarios combining different policies and
technical support levels, they identify paths to sustainable regional development. Their
results suggest that improving technical and political support is key to guaranteeing
sustainable development rather than further restricting urbanization.
In our work, we examine scenario planning to achieve the SDG goals in a rapidly
changing environment. Useful insights in this area are provided by Aguiar et al. [18],
who propose a novel approach to co-design global target-seeking scenarios by capturing
multiple and contrasting perspectives on pathways to achieving the SDGs. Recognizing
that implementing the SDGs requires coordinated actions across local, national, regional,
and global levels, they emphasize the importance of including diverse perspectives from
various scales and geographic regions. Their methodology involves a multi-stakeholder
process that captures global perspectives through a review of existing scenarios, gathers
sub-global perspectives, and analyzes the convergences and divergences between these
viewpoints. In their case study of the African Dialogue on The World in 2050, they uncover
divergent themes—such as urbanization, population growth, agricultural practices, and
the roles of different actors—that challenge the underlying assumptions of the existing
global sustainability scenarios, highlighting the necessity of incorporating local and re-
gional nuances into global scenario planning. Similarly, Wilts and Britz [19] underscore the
critical need for detailed disaggregated data to effectively assess progress toward the SDGs.
They develop an SDG indicator framework for dynamic Computable General Equilibrium
models, incorporating 68 endogenous indicators related to 15 SDGs with household-level
detail based on a micro-simulation. Applying this framework in a global analysis focusing
on selected low- and lower-middle-income countries, they find significant sustainability
gaps by 2030 and 2050, especially in the environmental domain. Their analysis reveals
that trade-offs exist among and within SDGs, as well as across different socioeconomic
pathways. Notably, they observe increasing inequality over time for several indicators,
regardless of the average aggregate household developments, pointing to the need for
targeted redistribution and compensation policies. Together, these studies emphasize that
achieving the SDGs requires not only capturing diverse perspectives in scenario plan-
ning but also incorporating detailed, disaggregated data to understand distributional
impacts and trade-offs. By integrating multiple scales of perspectives [18] and detailed
socioeconomic data [19], policymakers and researchers can develop more nuanced and ef-
fective strategies for sustainable development that account for inequalities and the complex
interplay between different SDGs.
Since our research focuses on the national level, Allen et al.’s [20] study that reviews
and assesses quantitative models that have the potential to support national develop-
ment planning for the SDGs is important for us. They develop a typology and inven-
tory of 80 different models and highlight gaps in the model capabilities for analyzing
all the SDG targets within a single framework. They suggest that combining top-down
macro-framework models with bottom-up sectoral models provides a robust approach for
analysis and decision-making, enabling policymakers to explore trade-offs and synergies
among sectors.
Academic institutions that have links with industrial companies can also significantly
contribute to sustainable development. In this field, Beynaghi et al. [21] focus on the role of
universities in advancing sustainable development. They systematically analyze the impli-
cations of the sustainable development trends and the future directions that universities
might take under a potential second decade of Education for Sustainable Development
(2015–2024). By projecting the current trends into the future, they frame possible future
orientations through three unique scenarios: a socially oriented university, an environmen-
tally oriented university, and an economically oriented university. Each scenario entails
fundamental changes affecting the university’s mission, focus areas, disciplines, view of
Education for Sustainable Development, external partners, projects, geographical focus,
and main functions involved. Their work provides conceptual and practical instruments for
scholars, university leaders, and policymakers to consider strategically how these futures
might be realized.
These studies collectively demonstrate the critical role of scenario planning in ad-
dressing sustainable development and disaster risk reduction. They highlight the im-
portance of integrating disaster considerations into development planning, involving
multiple stakeholders, and capturing diverse perspectives to create comprehensive and
effective scenarios.
2.3. Challenges and Limitations in Scenario Planning

Developing sustainable development scenarios that adequately account for disaster
impacts is a demanding process. In our view, several key challenges emerge when an
organization attempts to create scenarios for sustainable development:
1. Achieving a clear description of the long-term goals that the company must fulfill
within the planned period.
2. Collecting and assessing data on influential factors (drivers), determining their impact
magnitude and the trends of their change over time.
3. Selecting viable scenario options and estimating the likelihood of their successful
implementation.
4. Producing detailed scenario narratives that are actionable, including both potential
benefits and possible losses, to ensure their practical applicability.
5. Coordinating the scenarios with interested groups, ensuring that diverse stakehold-
ers—both internal and external—are engaged, informed, and supportive.
These steps collectively underscore the resource-intensive nature of scenario planning.
Such complexity often limits organizations with fewer resources or less expertise from
fully developing robust scenarios. The complexity of analyzing multiple SDG targets and
variables is evident from studies like that of Allen et al. [20], who highlight that the existing
models and approaches may not fully capture all the aspects of sustainable development
within a single framework. Similarly, research by Nieto-Romero et al. [22] illustrates that
information sharing and visioning alone do not guarantee change; scenarios must align with
longer-term agendas, address differing barriers to action, and reconcile trade-offs among
stakeholder groups. These findings reinforce the significance of considering stakeholder
involvement and integrating scenarios into broader strategic processes—points that align
closely with our emphasis on coordination and actionable scenario narratives.
Moreover, the methodological advancements discussed by Kiviluoto et al. [23] under-
score the need for refined analytical techniques, such as specialized Delphi approaches, to
systematically manage qualitative data and derive scenarios. This supports our assertion
that a thorough understanding of influential drivers and the robust evaluation of scenarios’
feasibility are crucial. Ultimately, addressing these challenges—clear goal-setting, com-
prehensive data gathering, viable option selection, detailed scenario construction, and
strong stakeholder coordination—can guide organizations toward more effective scenario
planning, even in complex and resource-constrained contexts.
2.4. Challenges and Limitations in Scenario Planning (Old)

Despite the recognized benefits, developing sustainable development scenarios that
adequately account for disaster impacts remains a labor-intensive and resource-demanding
process. Organizations, particularly in developing countries, often lack the necessary
resources and expertise to conduct thorough scenario analyses.
Allen et al. [20] emphasize that, while sophisticated models are available to sup-
port national development planning for the SDGs, gaps remain in their ability to ana-
lyze all the SDG targets and variables within a single modeling framework. They high-
light the need for combining different modeling approaches to capture the complexity of
sustainable development.
Nieto-Romero et al. [22] assess the opportunities and limitations of scenario planning
in shaping a tangible agenda for sustainable development within a rural community in
Transylvania. They find that, while scenario planning was useful for articulating a shared
development trajectory, the actors perceived different barriers to action, and the trade-
offs accepted for collaboration differed among the groups. Their study suggests that
information sharing and visioning alone are not sufficient to catalyze change, emphasizing
that scenarios should be integrated into longer-term agendas and strategies.
Kiviluoto et al. [23] conduct a Delphi study with 30 walking and cycling experts in
Finland to explore transport experts’ views on the future and derive scenarios of walking
and cycling within the broader context of the urban mobility system. They create five sce-
narios depicting walking and cycling in 2034, which can be used as a basis for strategic
transport planning and policy. Their methodological elaboration of the disaggregative
Delphi analysis systematizes the analysis of qualitative data, contributing to advancements
in scenario planning methodologies.
These challenges highlight the need for innovative approaches and tools to make scenario
planning more accessible and effective, especially for organizations with limited resources.
2.5. Innovative Approaches and the Role of Artificial Intelligence

Integrating innovative approaches, including artificial intelligence (AI), offers potential
solutions to the challenges in scenario planning. According to existing research [24], AI
tools can enhance predictive capabilities and facilitate complex data analysis; thus, we
believe they could potentially help to streamline the scenario development process.
For example, advanced analytical frameworks have been applied in related domains
to guide decision-making under complex conditions. Petrudi et al. [25] introduced a social
sustainability innovation framework to assess suppliers during the COVID-19 pandemic,
applying group grey best–worst and improved grey relational analysis methods. Their
findings, highlighting “safety and health practices”, “remote working conditions”, and
“localization” as the key criteria, demonstrate that robust analytical techniques can identify
essential factors influencing organizational resilience and sustainability. Although not
directly focused on disaster scenario planning, these methods underscore the value of
sophisticated analyses to prioritize actions and enhance long-term preparedness.
Similarly, studies on the financial implications of disasters illustrate the importance of
long-term perspectives. Montero et al. [26] analyze natural disasters’ impact on insurance
company volatilities using a multi-event framework and GARCH-type models, providing
policymakers and insurers with insights into volatility management and value-at-risk.
Although their work centers on financial markets, it exemplifies how thorough quantitative
analyses and model-based evaluations can inform strategic interventions that extend
beyond immediate disaster effects and support informed scenario development.
While direct application of AI in scenario planning is not extensively covered in the
provided literature, according to the authors of the current article, there is potential for AI
tools, such as large language models and conversational agents, to facilitate communication,
enhance data analysis, and support the generation of detailed scenarios. These technologies
might help organizations to overcome resource constraints by automating aspects of the
scenario development process and providing accessible platforms for stakeholder engage-
ment. Especially, it is important because, according to existing research [27], there is a need
to address the lack of control over governance that prioritizes financial performance and
the company’s contribution to a sustainable future.
2.6. Integrating Disaster Risk into Sustainable Development Scenarios

The increasing frequency and severity of disasters necessitate the integration of disaster
risk considerations into sustainable development scenarios. As highlighted by the studies
discussed, scenario planning that incorporates disaster impacts can help organizations and
policymakers to develop more robust and resilient strategies.
Kong et al. [28] propose a two-stage risk-neutral stochastic model to develop restora-
tion resource deployment and allocation strategies for critical infrastructure systems under
uncertain disaster scenarios. They apply a risk metric to assess the risk imposed by uncer-
tainty and present a multi-objective optimization model for risk management. Their results
demonstrate significant advantages in using risk-neutral strategies to improve system
resilience and assist infrastructure operators in selecting effective emergency response
strategies. One of the measures to reduce such risks is to increase transportation resilience
as a component of city sustainability [29].
These works underscore the critical need for integrating disaster risk management
into sustainable development planning and scenario analysis. By doing so, organizations
can enhance their resilience, better prepare for uncertainties, and contribute to achieving
the Sustainable Development Goals in a world where crises are increasingly frequent.
2.7. The Integration of Artificial Intelligence in Scenario Planning and Disaster Management
The integration of artificial intelligence (AI), particularly generative AI and conversa-
tional agents, presents a significant opportunity to enhance scenario planning and disaster
management practices. Traditional scenario planning often relies heavily on expert judg-
ment and may struggle to account for the complexity of future uncertainties [30]. The use
of AI can address some of these limitations by automating data analysis.
Hao et al. [30] explore the potential integration of AI techniques into scenario planning,
identifying applications in plan generation, scenario generation, and plan evaluation. They
argue that AI can think creatively and unconventionally, sometimes generating scenarios
beyond human imagination; to briefly justify this claim, we can refer to the famous example
of the “Move 37” made by AlphaGo Lee. AI algorithms can analyze vast amounts of data
to identify patterns and trends that may not be immediately apparent to human planners.
This capability can significantly expand the scope and depth of scenario planning, enabling
organizations to better prepare for adverse future scenarios.
Moreover, AI models such as deep learning and reinforcement learning excel at cap-
turing implicit and dynamic interactions within complex systems [30]. In the context of
disaster management, these models can simulate the impacts of various disaster scenarios
on infrastructure, supply chains, and communities. If organizations successfully incorpo-
rate AI into scenario planning and achieve high-quality results, they can enhance their
resilience by identifying vulnerabilities and testing the effectiveness of different strategies
within a range of conditions.
Cheng et al. [31] demonstrate the practical application of AI in disaster management
through a human–AI teaming workflow for estimating post-disaster debris volume and
composition. Their approach leverages drones for rapid data collection and utilizes AI
models alongside crowdsourcing to detect damaged buildings and assess the damage levels
based on aerial imagery. By connecting building the damage states to the generated debris,
they provide efficient and accurate debris estimations, reducing the time required from
days to hours and minimizing the predictive uncertainty by up to 40%. This case study
underscores how AI can support rapid decision-making and resource allocation in disaster
response efforts.
Furthermore, the use of conversational AI agents, such as ChatGPT, holds promise for en-
hancing knowledge dissemination and decision support in technical fields. Hostetter et al. [32]
evaluate the performance of ChatGPT and early version of Google’s Bard in handling
fire-safety-related queries. They find that ChatGPT demonstrates relatively superior per-
formance, providing detailed and correct responses to technical questions. While noting
limitations such as potential biases and inaccuracies, the authors suggest that, as chatbot
technology matures, it could revolutionize engineering practices by providing instant
access to critical information.
These existing advancements by other authors highlight the potential of AI, particu-
larly generative AI and conversational agents, to transform scenario planning and disaster
management. By leveraging AI’s capabilities, organizations could potentially overcome
resource constraints, improve the accuracy of their predictions, and enhance their capacity
to prepare for and respond to disasters while pursuing the Sustainable Development Goals.
2.8. Challenges and Limitations of ChatGPT Use for Scenario Generation

Using AI tools like ChatGPT for scenario generation also presents challenges and
limitations. Although it could potentially enable organizations with limited resources to
identify priority areas and enhance competencies, effective application requires careful
prompt engineering and ongoing verification of outputs. In our previous work [33], we
argued that successful application relies on careful prompt construction and iterative re-
finement of outputs. This process ensures that the generated scenarios remain coherent,
contextually relevant, and useful for addressing the complex challenges in sustainable
development and disaster management. Moreover, according to Spaniol et al. [34], while AI
can produce large volumes of scenarios at minimal costs, these outputs often lack the con-
textual richness and domain-specific insight that experienced scenario facilitators provide.
This underscores the importance of iterative refinement and expert oversight. Following
these practices should increase the efficiency of ChatGPT use in scenario generation. With-
out active human involvement to refine and adapt AI-generated narratives, the resulting
scenarios risk oversimplifying strategic uncertainties and failing to inspire meaningful
long-term planning. In other words, AI-driven scenario generation can support scenario
planning processes, but it must be coupled with human expertise, ethical considerations,
and a robust methodological framework to ensure high-quality, actionable outcomes.
3. The Methodology of the Research

In this section, we present the methodology of the research, i.e., how we address the
assessment of the effects of disasters on Sustainable Development Goals in companies. The
main research data were obtained via a survey.
3.1. Analyzing Dependencies: Correlation and Mutual Information Analyses

To analyze the dependencies between variables in our dataset, we employed two
complementary approaches: correlation analysis and mutual information (MI) analysis.
Each method captures different aspects of the relationships between variables, enabling
us to explore both linear and non-linear dependencies. Correlation coefficients are well-
established statistical measures for examining linear relationships [35], while MI, rooted in
information theory, quantifies the amount of shared information between variables without
assuming linearity [36].
3.1.1. Correlation Analysis

Correlation analysis involves calculating the Pearson correlation coefficient, which
measures the strength and direction of linear relationships between variables. The correla-
tion coefficient r between two variables X and Y is defined as
Cov( X, Y )
r= ,
σX σY
where Cov( X, Y ) is the covariance between X and Y, and σX and σY are the standard
deviations of X and Y, respectively.
A correlation matrix was constructed, with each element representing the correlation
coefficient between two variables. The resulting matrix is symmetric (rij = r ji ) and pro-
vides a straightforward way to assess pairwise linear dependencies. Variables with high
absolute correlation values (|r | > threshold) indicate a strong linear relationship. This
analysis is particularly useful for identifying variables that co-vary in a straightforward,
proportional manner.
The correlation matrix was visualized as a heatmap, where the color intensity corre-
sponds to the strength of the correlation.
3.1.2. Mutual Information with Null Model Thresholding

Mutual information (MI) quantifies the amount of shared information between two
variables, capturing both linear and non-linear dependencies [36]. For discrete variables,
MI is computed using

p( x, y)
MI( X, Y ) = ∑ ∑ p( x, y) log ,
x ∈X y∈Y
p( x ) p(y)
where p( x, y) is the joint probability mass function of X and Y, and p( x ) and p(y) are their
marginal probability mass functions.
Unlike correlation, MI does not assume linearity, making it sensitive to a broader range
of relationships. The resulting MI matrix is symmetric (MI( X, Y ) = MI(Y, X )) and provides
a general measure of dependency. However, interpreting MI values can be challenging as
the scale of MI depends on the entropy of the variables.
To address the lack of clear interpretability, we implemented a null model approach to
establish a significance threshold for MI values. This approach is less commonly used but
provides a robust statistical basis for interpreting MI.
Null Model Approach

The null model approach generates a baseline distribution of MI values under the
assumption of independence. For each pair of variables, one variable (X) is randomly
permuted while the other (Y) is kept unchanged. This process simulates the absence of any
true dependency between the variables. The steps are as follows:
1. For each pair of variables X and Y, permute X 100 times, creating random pairings
with fixed Y.
2. Compute the MI for each permutation using the formula for mutual information.
3. Aggregate these MI values to form a null distribution.
4. Define the significance threshold as the 99th percentile of the null distribution. MI
values above this threshold are considered to be statistically significant.
By applying this approach to every pair of variables in the dataset, we derived pair-
specific thresholds that reflect the unique statistical properties of the data. The MI matrix
was visualized as a heatmap, with significant dependencies highlighted.
3.1.3. Comparison of Approaches

While correlation analysis is well-suited for identifying linear dependencies, mutual
information (MI) regression provides a more general framework for detecting both linear
and non-linear relationships. Together, these methods provide a comprehensive view of
the relationships between variables, facilitating a deeper understanding of the underlying
structure in the data.
An advantage of correlation analysis is its ability to indicate the direction of linear
relationships through the sign of the correlation coefficients. Negative values in the correla-
tion matrix reveal inverse relationships between variables, enabling more nuanced insights.
For instance, a negative correlation coefficient between two variables indicates that, as one
variable increases, the other tends to decrease. This directional information is valuable for
understanding trade-offs or compensatory mechanisms within the system.
In contrast, MI is a non-negative measure that quantifies the amount of shared infor-
mation between variables but does not convey the direction of the relationship. While MI
effectively captures both linear and non-linear dependencies, it treats positive and negative
associations equivalently. Therefore, combining correlation analysis with MI enables us to
identify not only the presence and strength of dependencies but also their directionality.
Both the correlation and MI matrices were visualized as heatmaps. Hierarchical
clustering using Ward’s method [37] was applied to reorder the variables based on the
MI values, enabling patterns of dependency to emerge visually. The inclusion of negative
correlation values in the heatmap adds depth to the analysis by highlighting inverse
relationships, which might be overlooked if only MI values were considered.
3.2. Factors for Evaluation

Table 1 summarizes the key factors used for evaluation in this research. These factors
have been selected based on their common use and relevance in studies by other authors,
ensuring that they align with widely accepted parameters in the scientific community. Each
factor is listed alongside references to studies where similar parameters were analyzed,
highlighting their importance and application in various contexts. This comprehensive
overview serves as the basis for the evaluation in this research.
Table 1. Factors and corresponding references.
The Authors Who Also Studied

No. Parameter
Similar Parameters
B1.1 Creativity of company employees [38]
B1.2 Entrepreneurship of company employees [38]
B2.1–B2.17 Sustainable Development Goals (SDGs) [15,16,39]
B3.1 Risk management and increasing disaster resilience [28,40–44]
B3.2 Rapid response and liquidation of disaster consequences [28,42]
B3.3 The use of artificial intelligence in disaster planning and management [40–43]
B3.4 Attracting and mobilizing stakeholders [38,45]
B3.5 Securing supply chains [28]
B3.6 Ensuring sustainability in the event of disasters [42,46,47]
B3.7 Temporary accommodation of residents [38,46,48]
B3.8 Data storage and protection [44]
C1.1 Disconnection of electricity [28,45,47,49]
C1.2 Water supply interruption [28,45,47,49]
C1.3 Interruption of food supply [50]
C1.4 Closed roads [45,49]
Table 1. Cont.
No. Parameter The Authors Who Also Studied Similar Parameters

C1.5 Disconnection of communication [48]
C1.6 Destruction in the company [38]
C1.7 Fire in the company [40]
C1.8 A flood in the company [40,49]
C1.9 Absence of heating and ventilation [51]
C1.10 Problems with work organization due to the pandemic [38,52]
C1.11 Problems with suppliers [38,45]
C1.12 Problems with customers [38,45,52]
C1.13 Injuries to employees [45,47]
C1.14 Deaths of employees [45,47]
C1.15 Information system violations due to hacking and viruses [53]
C1.16 Psychological stress of employees [45,47]
C1.17 Legal disputes [54]
C1.18 Damage to reputation [55]
C1.19 Financial losses [38,47,54]
C1.20 Human resource issues [45,47,49]
3.3. Methodology for Scenario Generation

To demonstrate the potential of generative AI in scenario planning, we utilized Ope-
nAI’s ChatGPT, a conversational AI model based on the GPT-4 architecture, to generate
scenarios related to sustainable development and disaster management. The objective was
to demonstrate how AI-generated scenarios can be employed in organizational planning
and disaster preparedness efforts.
3.3.1. Prompt Engineering Approach

Prompt engineering is the process of crafting input prompts to guide AI models like
ChatGPT to produce desired outputs. Effective prompt engineering is critical to ensure that
the generated scenarios are relevant, coherent, and of high quality [56]. Although we do
not pretend to be novel here, in our approach, we applied prompt engineering techniques
that enabled us to obtain the desired quality of the output. This involved the following:
• Defining Clear Objectives: Explicitly stating the purpose of the scenario generation,
focusing on specific themes such as Sustainable Development Goals (SDGs), disaster
risk management, and organizational resilience.
• Providing Contextual Information: Including relevant background information in the
prompts, such as the organizational context, industry sector, and specific challenges
faced. This helps ChatGPT to tailor the scenarios to our needs; in our prompts, we
included the knowledge extracted from conducted survey.
• Specifying Desired Output Structure: Guiding the AI to produce scenarios with a clear
structure, including elements like timeframes, key actors, and potential outcomes.
This ensures that the scenarios are comprehensive and actionable.
• Incorporating Key Variables and Constraints: Highlighting critical factors and uncer-
tainties to be considered in the scenarios, such as technological advancements, climate
change impacts, and regulatory environments.
• Iterative Refinement: Engaging in an iterative process, reviewing the AI’s outputs,
and refining the prompts as needed to enhance clarity and relevance.
3.3.2. Approach to AI-Generated Scenarios

To increase the quality and relevance of the AI-generated scenarios, we took the
following measures:
• Ethical Considerations: Followed ethical guidelines, ensuring that the prompts and
generated content respected confidentiality, avoided biases, and complied with appli-
cable policies [57].
• Preliminary Review: Conducted a preliminary review of the AI-generated scenarios
for plausibility and alignment with the prompts. This involved checking for logical
consistency and relevance to the specified context.
• Documentation: Documented the prompt engineering process, including iterations
and refinements, to maintain transparency and replicability. However, regarding the
current state of GenAI technology, the replicability property is quite limited.
By employing all these techniques, we aimed to ensure that the scenario generation
process via ChatGPT yielded results meeting our quality expectations. More details will be
provided at the end of the next section presenting the results.
4. Results
The study involved 30 respondents from Lithuanian companies. The companies’
requests were selected focusing primarily on large businesses. The survey respondents
who were selected as specialists are familiar with making strategic decisions within the
company or have an opinion on strategic issues. Part of the survey was distributed via
email. Around one-third of the candidates responded to the requests. The other data
were collected from participants during university-organized events for manufacturing
and/or engineering companies. The majority of the respondents were engineers and
designers (37%), followed by directors, department heads, and managers (33%). The
rest were scientists (7%), as well as salespeople, marketing staff, and employees from
other specialties (23%). Most of the respondents (90%) had higher education qualifica-
tions. We provide mean values along with standard deviation values for all the responses
in Appendix C.
Regarding the size of the companies, half of the respondents (50%) were from large
companies, 23% from small- and medium-sized companies (SMEs), and 27% from micro-
companies. The primary activities of the surveyed companies were manufacturing (60%),
while the rest specialized in construction, transport, sales, science, and fishing. There-
fore, the results of our study are more reflective of large companies, particularly in the
manufacturing sector.
When asked whether they had encountered issues regarding sustainable development
and disaster management, 37% of the respondents indicated that they had, 47% said they
had not, and the remaining 16% had no opinion on the matter. Similarly, 37% of the
respondents indicated that their company develops sustainable development and disaster
response scenarios (i.e., there are official documents declaring this within the company),
33% said their company does not, and the rest had no opinion.
According to the respondents’ answers, it is rare for companies to proactively en-
gage in sustainable development and disaster prevention initiatives on their own. While
external stakeholders like the state, suppliers, consumers, and other interested parties
often participate in these processes, 57% of the respondents indicated that, in practice,
only the company itself is involved in addressing issues of sustainable development and
disaster management.
Regarding disaster preparedness, representatives of large companies rated their com-
panies higher (3.5 points out of 5 possible) compared to representatives of small- and
medium-sized companies (2.9 points) and micro-companies (2.8 points). The responses
were rated on a scale from 1 (very bad) to 5 (very good).
The conducted studies showed that respondents from large companies assess their
organizations’ preparedness for disasters better than representatives of small- and medium-
sized or micro-companies. Some differences are also visible (see Figure 1) in the assessment
of the SDGs in the context of disasters. For example, the representatives of micro-companies
assessed the importance of gender equality, climate action, industry, innovation, and infras-
tructure nearly one point lower. The majority of the respondents from all the companies
surveyed agreed that one of the most important SDGs is clean water and sanitation. They
similarly assessed good health and well-being, the absence of hunger, life on land and
below water, and responsible consumption as important factors.
The average ratings of the suitability of possible scenario types for disaster manage-
ment were similar in large-, small-, and medium-sized companies. In micro-companies,
these ratings were somewhat lower. “Rapid response and liquidation of disaster conse-
quences” was rated as the most important scenario. Other highly evaluated scenarios
included “securing supply chains”, “attracting and mobilizing stakeholders”, and “tempo-
rary accommodation of residents” (see Figure 2).
Micro companies (less than 10 employees)

Small and medium-sized companies (employs between 10 and 249 persons)
Large companies (employs 250 or more persons)
B2.1 - No poverty
B2.17 - Partnerships for the 5 B2.2 - Zero hunger
goals
4.5
B2.16 - Peace, justice & 4 B2.3 - Good health and well-
strong institutions being
3.5
3
2.5
B2.15 - Life on land B2.4 - Quality education
2
1.5
1
B2.14 - Life below water 0.5 B2.5 - Gender equality
0
B2.6 - Clean water and

B2.13 - Climate action
sanitation
B2.12 - Responsible B2.7 - Affordable and clean

consumption & production energy
B2.11 - Sustainable cities and B2.8 - Decent work and

communities economic growth
B2.9 - Industry, innovation,
B2.10 - Reduced inequalities
and infrastructure
Figure 1. The importance of different SDGs for sustainable development in the context of disaster
management.
In the surveyed companies, the respondents mostly experienced negative impacts of

disasters, such as disconnection of electricity, disconnection of communication, problems
with suppliers, problems with customers, challenges with work organization due to the
pandemic, financial losses, human resource issues, and psychological stress among employ-
ees. These impacts were assessed on average between 2.3 and 2.7 points on a scale from 1
to 5. Other impacts of disasters occurred even less frequently (see Figure 3).

B3.1 - Risk management and

increasing of disaster resilience
5
4.5
4 B3.2 - Rapid response and
B3.8 - Data storage and
liquidation of disaster
protection 3.5
consequences
3
2.5
2
1.5
1
0.5 B3.3 - The use of artificial
B3.7 - Temporary
0 intelligence in disaster planning
accommodation of residents
and management
B3.6 - Ensuring sustainability in B3.4 - Attracting and mobilizing

the event of disasters stakeholders
B3.5 - Securing supply chains
Figure 2. The estimation of the relevance of possible scenario types for managing catastrophes.

C1.1 - Disconnection of electricity

C1.20 - Human resources issues 5 C1.2 - Water supply interruption
4.5
C1.19 - Financial losses C1.3 - Interruption of food supply
4
3.5
C1.18 - Damage to reputation 3 C1.4 - Closed roads
2.5
2
C1.5 - Disconnection of
C1.17 - Legal disputes 1.5 communication
1
0.5
C1.16 - Psychological stress of
0 C1.6 - Destruction in the company
employees
C1.15 - Information system

violations due to hacking and C1.7 - Fire in the company
viruses
C1.14 - Deaths of employees C1.8 - A flood in the company
C1.9 - Absence of heating and

C1.13 - Injuries to employees
ventilation
C1.10 - Problems with work
C1.12 - Problems with customers
organization due to the pandemic
C1.11 - Problems with suppliers
Figure 3. The frequency of negative impacts during catastrophes in companies.
4.1. Dependence Between Score Variables

Figure 4 shows the heatmap of the correlations between the scores for various ques-
tions. Only those questions with at least one correlation coefficient greater than 0.7 are
included here, while the full heatmap is provided in Appendix B (Figure A1).
C1.19 - Financial losses 1.0

C1.6 - Destruction in the company
C1.14 - Deaths of employees
C1.16 - Psychological stress of employees 0.8
C1.17 - Legal disputes
C1.18 - Damage to reputation
B2.14 - Life below water
B2.15 - Life on land 0.6
B2.11 - Sustainable cities and communities
B2.7 - Affordable and clean energy
B2.12 - Responsible consumption & production 0.4
B2.8 - Decent work and economic growth
B2.9 - Industry, innovation, and infrastructure
B2.17 - Partnerships for the goals
B2.1 - No poverty 0.2
B2.16 - Peace, justice & strong institutions
B2.3 - Good health and well-being
B2.6 - Clean water and sanitation 0.0
B3.2 - Rapid response and liquidation of disaster consequences
B3.8 - Data storage and protection
B3.4 - Attracting and mobilizing stakeholders
B3.1 - Risk management and increasing of disaster resilience 0.2
C1 1
C1 6
C1 1
C1 1
C1 1
B2 1
B2 1
B2 1
B2 1
B2 7
B2 1
B2 8
B2 9
B2 1
B2 1
B2 1
B2 3
B3 6
B3 2
B3 8
B3 4
B3 1
.5
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
C1
Figure 4. Partial heatmap of correlations between scores for different questions.
As shown in Figure 4, one cluster identified via hierarchical clustering (using Ward’s
method) includes the questions from group C: financial losses, destruction in the company,
death of employees, psychological stress of employees, legal disputes, and damage to
reputation. This indicates a tendency among the respondents to evaluate these questions
similarly, either increasing or decreasing their scores collectively. Notably, “psychological
stress of employees” was the least correlated with the other items in this cluster.
Additionally, the fourth column in the heatmap shows a negative correlation between
the scores from C1.16 psychological stress of employees and the scores from group B, which
includes the SDGs’ importance scores. This suggests that those respondents who reported
more frequent occurrences of psychological stress tended to rate sustainability factors
as less important during disasters. This finding highlights the potential significance of
stress-reducing measures in supporting sustainability-related initiatives. We hypothesize
that psychological challenges may reduce employees’ motivation to focus on long-term
goals, such as achieving SDGs.
Figure 5 presents the heatmap of the mutual information (MI) scores. The questions
were filtered using the null model approach described in Section 3.1.2. The significance
threshold for the 0.99 percentile was estimated as 0.467, and only questions with at least
one MI score exceeding this value are shown.
Unlike linear correlation coefficients, MI highlights logical dependencies, such as
the relationships between assessments of equality-related SDGs (e.g., B2.10 and B2.5) or
between assessments of “Life below water” and “Life on land” SDGs (B2.14 and B2.15,
respectively). Questions B3.1 and B3.5 are strongly related to each other in both the
correlation and MI matrices. We observed that the score for “Securing supply chains” (B3.5)
is a very good predictor, showing high MI scores with B3.1, B3.4, and B3.8.
There could be a potential informational overlap when we consider questions B3.5
and B3.1 as the mutual information between B3.5 and B3.1 is very high. Therefore, we
could assume that we might leave out either question B3.5 or B3.1 in the survey without
significant loss of information since they provide similar insights. However, since the
mutual information between B3.1 and B3.8 is much lower (slightly above 0.99 percentile
significance threshold 4.67) than between 3.5 and 3.8, the question score from B3.1 offers
additional valuable information. This suggests that we should retain question B3.1 in the
survey as it captures unique aspects that are not fully covered by B3.5 or B3.8.
0.8
0.7
B3.4 - Attracting and
mobilizing stakeholders
0.6
0.5
B3.8 - Data storage
and protection
0.4
B3.1 - Risk management and
increasing of disaster resilience
0.3
0.2
B2.5 - Gender equality
0.1
B2.15 - Life on land
0.0
.5
.4
.1
.8
.1
.1
.5
.1
B3
B3
B2
B3
B3
B2
B2
B2
Figure 5. Heatmap of mutual information scores.
4.2. Scenario Generation Using ChatGPT

As discussed in Section 3.3.2, in this section, we will showcase the scenario generation
using state-of-the-art generative AI technology, more specifically ChatGPT by OpenAI
(Sanfrancisco, CA, USA), the latest version, o1-preview, which is based on ChatGPT 4.0.
4.2.1. Information Provided in Scenario Prompt Generation

While generating scenarios using conversational AI, to achieve our subjective assess-
ment regarding high-quality scenarios, it was essential to include specific information in
the prompts [58]:
1. Clear Scenario Focus: Define the main topic or challenge the scenario should ad-
dress. For example, “Develop a scenario where a manufacturing company adapts to
increasing climate-related disasters over the next decade”.
2. Contextual Background: Provide background information about the organization, indus-
try, or region. This may include size, location, key stakeholders, and current strategies.
3. Timeframe: Specify the temporal scope of the scenario, e.g., “over the next 10 years”
or “by 2030”.
4. Key Drivers and Uncertainties: Identify critical factors that influence the scenario, such
as technological developments, economic trends, policy changes, or environmental
challenges.
5. Desired Level of Detail: Indicate the expected depth and complexity of the scenario.
For example, request a detailed narrative that includes specific actions, consequences,
and stakeholder responses.
6. Constraints and Assumptions: Mention any constraints or assumptions to be consid-
ered, such as resource limitations, regulatory frameworks, or social dynamics.
7. Output Format: Specify if the scenario should be presented in a particular format,
such as a narrative, bullet points, or a structured outline.
An example of a prompt fulfilling the requirements above might be the following:
“Imagine a scenario where a large manufacturing company in Lithuania faces
an increasing frequency of natural disasters due to climate change over the next
decade. The company aims to enhance its disaster preparedness and sustain-
ability practices while maintaining profitability, despite resource limitations and
stringent environmental regulations. The company must also address stakeholder
concerns, including those of employees, local communities, and government agen-

cies. Provide a detailed description of how the company adapts its operations,
engages with stakeholders, and leverages artificial intelligence—specifically ma-
chine learning and predictive analytics—in disaster planning and management.
Include potential challenges, ethical considerations, and outcomes. Present the
scenario in a structured format with clear headings for each section, such as
Introduction, Adaptive Strategies, Stakeholder Engagement, AI Implementation,
Challenges, Ethical Considerations, and Outcomes.”
Here is a breakdown of the prompt:
1. Clear Scenario Focus:
• “Imagine a scenario where a large manufacturing company in Lithuania faces an increas-
ing frequency of natural disasters due to climate change over the next decade.”
2. Contextual Background:
• “A large manufacturing company in Lithuania...”
3. Timeframe:
• “...over the next decade.”
4. Key Drivers and Uncertainties:
• “Increasing frequency of natural disasters due to climate change.”
5. Desired Level of Detail:
• “Provide a detailed description of how the company adapts its operations, engages with
stakeholders, and leverages artificial intelligence—specifically machine learning and
predictive analytics—in disaster planning and management. Include potential challenges,
ethical considerations, and outcomes.”
6. Constraints and Assumptions:
• “...while maintaining profitability, despite resource limitations and stringent environ-
mental regulations.”
7. Output Format:
• “Present the scenario in a structured format with clear headings for each section, such
as Introduction, Adaptive Strategies, Stakeholder Engagement, AI Implementation,
Challenges, Ethical Considerations, and Outcomes.”
4.2.2. Potential for Assessing AI-Generated Scenarios

Assessing the quality of AI-generated scenarios is important for understanding their
utility in organizational planning. While assessment was beyond the scope of our current
research, we acknowledge that future studies could benefit from evaluating AI-generated
scenarios based on established criteria for high-quality scenarios, such as plausibility,
relevance, internal consistency, differentiation, and comprehensiveness [59].
The key points for assessing AI-generated scenarios include the following:
• Plausibility: Determining whether the scenario is believable and grounded in
logical assumptions.
• Relevance: Ensuring that the scenario addresses the specific issues and objectives
relevant to the organization.
• Internal Consistency: Checking for logical coherence within the scenario
without contradictions.
• Differentiation: Comparing multiple scenarios to ensure that they explore distinct
pathways and outcomes.
• Actionability: Evaluating whether the scenario provides practical insights that can
inform strategic decision-making.
Future research could involve experts in sustainable development and disaster man-
agement to rigorously assess AI-generated scenarios using these criteria. This would
contribute to validating the effectiveness of AI in scenario planning and identifying areas

for improvement.
4.2.3. Implications for Scenario Planning Using Conversational AI

The use of conversational AI like ChatGPT in scenario planning offers significant
benefits, including rapid generation of diverse scenarios, enhanced creativity, and the
ability to incorporate complex variables. However, the quality of the output is highly
dependent on the quality of the input prompts.
Effective prompt engineering is crucial for guiding the AI to produce valuable sce-
narios. By providing clear objectives, detailed context, and specific instructions, organi-
zations can leverage AI to support strategic planning and enhance their preparedness for
future uncertainties.
While our study demonstrates the potential of AI in generating scenarios, assessing
the quality and utility of these scenarios remains an important area for future research.
By exploring assessment methodologies and involving domain experts, organizations can
better understand how to integrate AI-generated scenarios into their planning processes.
4.3. ChatGPT-Generated Scenarios

In this section, we provide the ChatGPT-generated scenarios according to the prompts
in Appendix A. In the conversation, we also demonstrate how the data of our study findings
were reflected in the generated scenarios by ChatGPT.
4.3.1. Pessimistic Scenario (2025–2035)

Company Profile
Baltic Car Makers is a large Lithuanian manufacturing company specializing in auto-
motive parts production. As a key player in the national economy, it employs a significant
number of professionals, including engineers, designers, and managers.
Initial Challenges
Despite its size, the company has limited prior engagement in sustainable development
and disaster management practices. With climate change accelerating, natural disasters
become more frequent, directly impacting operations. The existing vulnerabilities include
the following:
• Frequent electricity and communication disruptions.
• Supply chain issues with both suppliers and customers.
• Organizational challenges during pandemics.
• Psychological stress among employees.
Impact of Possible Disasters

Between 2025 and 2035, Lithuania experiences an increase in severe natural disasters:
• High Probability: Floods, storms, and heatwaves become more common due to
climate change.
• Size and Extent: Disasters affect large regions, disrupting national infrastructure.
• Effects on the Company:
– Operations are frequently halted.
– Supply chains are severely disrupted.
– Physical infrastructure suffers significant damage.
– Employee well-being deteriorates due to constant crises.
Actions to Increase Organizational Resilience

Baltic Car Makers makes minimal efforts to enhance disaster preparedness:
• Internal Efforts: Reluctance to invest in sustainability due to financial constraints.
• Lack of Initiative: Waits for external stakeholders to lead, but such support is limited.
• Failure to Collaborate: Does not actively seek partnerships or government assistance.
Stakeholder Actions
Stakeholder engagement is weak:
• Employees: High stress levels lead to decreased productivity and increased turnover.
• Suppliers and Customers: Face their own challenges, causing further supply
chain breakdowns.
• Government Agencies: Provide minimal support due to resource limitations.
Integration of Artificial Intelligence

The company hesitates to invest in AI technologies:
• Lack of Adoption: No implementation of AI for disaster planning or supply
chain management.
• Technological Lag: Competitors adopt AI, leaving Baltic Car Makers behind.
Addressing Sustainable Development Goals

Minimal progress is made on SDGs:
• Important SDGs Neglected: Little action on clean water (SDG 6), good health (SDG 3),
or responsible consumption (SDG 12).
• Less-Prioritized SDGs Ignored: Gender equality (SDG 5), climate action (SDG 13), and
innovation (SDG 9) remain unaddressed.
Resilience Enhancements
The company fails to strengthen its resilience:
• Infrastructure: No significant upgrades; facilities remain vulnerable.
• Policies and Training: Lacks disaster preparedness policies; no employee training.
Overcoming Barriers
Financial and operational barriers persist:
• Financial Constraints: No efforts to secure funding or reallocate resources.
• Negative Impacts Unmitigated: Disasters continue to cause significant operational losses.
Outcomes and Impact

Negative outcomes prevail:
• Disaster Preparedness Ratings: Remain low; company is ill-prepared for crises.
• Operational Continuity: Frequent disruptions lead to declining productivity
and profitability.
• Competitive Advantage: Loses market share to better-prepared competitors.
Future Outlook
The long-term outlook is bleak:
• Survival: Company struggles to survive amidst mounting challenges.
• Reputation: Damaged due to poor crisis management and sustainability neglect.
• Missed Opportunities: Fails to capitalize on advancements in sustainable practices.
4.3.2. Most Likely Scenario (2025–2035)

Company Profile
Baltic Car Makers maintains its position as a leading automotive parts manufacturer
in Lithuania, contributing significantly to the economy and employment.
Initial Challenges
Recognizing the increasing frequency of natural disasters, the company begins to
engage in sustainable development and disaster management practices, albeit cautiously.

Natural disasters occur with moderate frequency:
• Medium Probability: Occasional floods and storms impact operations.
– Operations are temporarily disrupted.
– Supply chains experience delays but not complete breakdowns.
– Employee well-being is affected but managed.

The company takes moderate steps:
• Internal Efforts:
– Forms a basic disaster response team.
– Allocates limited funds for infrastructure improvements.
• Collaboration:
– Engages with suppliers to strengthen supply chains.
– Participates in industry forums on disaster preparedness.
Stakeholder Actions
Stakeholder engagement improves:
• Employees: Receive some training; morale stabilizes.
• Suppliers and Customers: Work together to mitigate disruptions.
• Government Agencies: Provide occasional support, although bureaucratic hurdles persist.

Partial adoption of AI technologies:
• Implementation:
– Introduces AI for basic supply chain management.
– Uses predictive analytics for maintenance scheduling.
• Benefits: Some operational efficiencies are realized.

Moderate progress is made:
• Important SDGs:
– Implements water-saving measures (SDG 6).
– Enhances workplace safety (SDG 3).
• Less-Prioritized SDGs:
– Begins to explore climate action initiatives (SDG 13).
– Invests modestly in innovation (SDG 9).
Strengthens resilience to a moderate extent:
• Infrastructure: Upgrades critical systems to withstand common disasters.
• Policies and Training: Implements basic disaster preparedness policies.
Overcoming Barriers
Some barriers are addressed:
• Financial Constraints: Secures modest funding through partnerships.
• Negative Impacts: Reduces operational disruptions through improved planning.
Outcomes and Impact

Mixed outcomes are observed:
• Disaster Preparedness Ratings: Improve to an average level.
• Operational Continuity: Stability is achieved, although not optimal.
• Competitive Advantage: Maintains market position but does not significantly excel.
Future Outlook
A cautious optimism prevails:
• Sustainability: Recognizes the need for ongoing efforts.
• Leadership Position: Viewed as a responsible company, although not a leader.
• Lessons Learned: Understands the importance of proactive measures.
4.3.3. Optimistic Scenario (2025–2035)

Company Profile
Baltic Car Makers emerges as a leader in sustainable manufacturing and disaster
preparedness, setting industry benchmarks in Lithuania and beyond.
Initial Challenges
Proactively addresses sustainable development and disaster management, transform-
ing challenges into opportunities.

Natural disasters have minimal impact due to effective strategies:
• High Probability: Disasters occur but are mitigated effectively.
– Operations remain uninterrupted.
– Supply chains are resilient and adaptive.
– Employee well-being is prioritized and maintained.

Significant initiatives are undertaken:
• Internal Efforts:
– Establishes a Sustainability and Resilience Department.
– Integrates SDGs into corporate strategy.
• Investment:
– Allocates 10% of annual revenue to sustainability initiatives.
– Invests in renewable energy, reducing dependence on external grids.
• Collaboration:
– Partners with universities for research on sustainable materials.
– Engages with international organizations for best practices.
Stakeholder Actions
Strong engagement leads to collective success:
• Employees: Highly engaged; benefit from wellness programs and training.
• Suppliers and Customers: Integrated into disaster preparedness plans.
• Government Agencies: Provide support and recognize the company’s leadership.

Advanced AI technologies are fully implemented:
• Implementation:
– Uses AI for predictive analytics and early warning systems.
– Optimizes supply chains and real-time decision-making during disasters.
• Benefits: Achieves high efficiency and swift response to incidents.

Significant progress across all the SDGs:
• SDG 6 (Clean Water): Implements water recycling in manufacturing.
• SDG 3 (Good Health): Offers comprehensive health programs.
• SDG 12 (Responsible Consumption): Adopts circular economy principles.
• SDG 5 (Gender Equality): Increases female representation in leadership.
• SDG 13 (Climate Action): Achieves carbon neutrality by 2030.
• SDG 9 (Industry, Innovation): Invests in cutting-edge technologies.
Company’s resilience is greatly enhanced:
• Infrastructure: Builds disaster-resistant facilities with renewable energy sources.
• Policies and Training: Establishes comprehensive policies; employees are well-prepared.
Overcoming Barriers
Effectively overcomes all barriers:
• Financial Constraints:
– Secures funding through green bonds and EU grants.
– Leverages investments and partnerships.
• Negative Impacts:
– Implements robust cybersecurity measures.
– Develops alternative supply chain strategies.
Outcomes and Impact

Positive outcomes are achieved:
• Disaster Preparedness Ratings: Ranks in the top 5% in Europe.
• Operational Continuity: Maintains uninterrupted operations.
• Competitive Advantage: Increases market share by 25%.
• Community Impact: Supports local communities through sustainable projects.
Future Outlook
The company is well-positioned for the future:
• Leadership: Receives international recognition for sustainability.
• Innovation: Continues to innovate with sustainable materials.
• Growth: Expands into new markets and sectors.
5. Discussion
The results of our survey provide valuable insights into how organizations perceive
and prioritize various factors related to sustainable development and disaster risk man-
agement. In this section, we discuss our findings in relation to the existing literature,
highlighting areas of convergence and divergence.
5.1. Entrepreneurship and Company Size

Our analysis did not reveal a correlation between the entrepreneurship of company
employees and the size of the company (the coefficient is 0.167). Meanwhile, the correlation
between company size and employee creativity was visible, 0.527. This suggests that smaller
companies should foster more employee creativity compared to larger organizations. One
possible explanation is that smaller companies offer more flexibility and opportunities for
employees to engage in innovative activities, whereas larger organizations may have more
resources for entrepreneurial initiatives.
This finding aligns with the observations of Sætra [60], who argues that businesses
often overlook critical sustainability dimensions such as social equity and inclusive eco-
nomic growth in their strategies. They emphasize the need for companies to adopt more
comprehensive approaches to sustainability that promote innovation and engage employ-
ees in the development of sustainable futures. In larger companies, the tendency to adhere
to established procedures may inhibit employee entrepreneurship, whereas smaller firms
may be more agile and open to innovative ideas.
5.2. Risk Management and Disaster Resilience

We found strong positive correlations between risk management and increasing disas-
ter resilience and several factors: the use of artificial intelligence in disaster planning and
management, attracting and mobilizing stakeholders, securing supply chains, and ensuring
sustainability in the event of disasters. This indicates that organizations that prioritize risk
management are also more likely to adopt advanced technologies, engage stakeholders,
secure their supply chains, and focus on sustainability during disasters.
This finding is consistent with the literature emphasizing the integration of AI and
stakeholder engagement in enhancing disaster resilience. For instance, Heo et al. [61]
discuss the feasibility of decarbonizing mega-scale industrial parks using AI-driven models
to integrate renewable energy sources and improve sustainability. Their study demon-
strates how AI can assist in planning and managing energy systems to mitigate climate
change impacts.
Similarly, Rohde et al. [62] present the Sustainability Criteria and Indicators for Artifi-
cial Intelligence Systems (SCAIS) Framework, highlighting the importance of considering
the multidimensional impacts of AI on sustainability. They advocate for a holistic approach
to AI implementation that addresses the environmental, social, and economic dimensions,
which aligns with our finding that AI use correlates with improved disaster resilience.
Moreover, Bag et al. [63] examine how multinational company (MNE) customer pres-
sures influence suppliers’ compliance and commitment to climate change adaptation (CCA)
and disaster risk reduction (DRR) goals. They find that MNE pressures can motivate suppli-
ers to develop dynamic capabilities and enhance sustainability performance, underscoring
the role of stakeholder engagement in disaster resilience.
5.3. Prioritization of Sustainable Development Goals

Our respondents rated clean water and sanitation, zero hunger, and good health
and well-being as very important factors. In contrast, factors related to climate control
and gender equality were evaluated as less important, with gender equality receiving a
significantly lower score than the other factors.
This prioritization reflects a focus on immediate and tangible needs over broader
and longer-term goals. Sætra [60] critiques the current business practices for emphasizing
technical aspects of sustainability while overlooking the critical dimensions, such as social
equity and inclusive growth. The low importance assigned to gender equality in our
survey suggests that organizations may not fully recognize its significance in achieving
sustainable development.
Additionally, Bag et al. [63] highlight that only MNE customer pressures effectively
motivate suppliers to comply with the CCA and DRR goals, implying that, without external
pressure, organizations may not prioritize climate action. This aligns with our finding that
climate control factors were considered to be less important by the respondents.
Cumming et al. [64] discuss the need for AI-ethics-led sustainability frameworks
that address issues like fairness, justice, and inclusiveness. The low prioritization of
gender equality in our survey indicates a gap in addressing these ethical considerations
within organizations.
5.4. Common Disaster Experiences

The most frequent disaster cases reported by our respondents were disconnection
of electricity, disconnection of communication, problems with customers, and psycholog-
ical stress of employees. These findings highlight the operational and human resource
challenges that organizations face during disasters.
Kim et al. [65] emphasize the importance of cultivating the disaster safety industry to
enhance disaster resilience and competitiveness. They note that government intervention
is often required to promote the disaster safety industry, suggesting that organizations may
struggle to address these challenges independently.
Furthermore, Jiang et al. [66] propose an agent-based model to assess the economic
ripple effects of disasters while considering firms’ adaptive behaviors. Their findings
demonstrate that firms’ adaptive behaviors can significantly reduce disaster losses. The
frequent occurrence of operational disruptions and employee stress reported in our sur-
vey underscores the need for organizations to develop adaptive strategies and enhance
their resilience.
5.5. Implications for Scenario Development and Organizational Resilience

Our findings have important implications for developing sustainable development sce-
narios that account for disaster consequences. The correlations between risk management
practices and factors such as AI use, stakeholder engagement, and supply chain security
suggest that integrating these elements into scenario planning can enhance
organizational resilience.
The low prioritization of climate control and gender equality indicates potential blind
spots in organizational strategies. As Sætra [60] and Cumming et al. [64] suggest, a more
comprehensive approach to sustainability that includes social and ethical dimensions is
necessary for achieving long-term sustainability goals.
The frequent occurrence of operational disruptions highlights the importance of ad-
dressing infrastructure vulnerabilities and human factors in disaster preparedness. Organi-
zations may benefit from investing in robust communication and power systems as well as
employee well-being programs to mitigate the impact of disasters.
5.6. Implications of AI Integration for Organizational Resilience and Sustainability

The integration of AI into scenario planning and disaster management aligns with
the literature emphasizing its potential to enhance organizational resilience. Hao et al. [30]
highlight how AI can automate plan generation, simulate complex scenarios, and evaluate
the performance of different plans under varying conditions. By utilizing AI-driven tools,
organizations can streamline the development of sustainable development scenarios that
account for disaster impacts without the extensive resources traditionally required.
However, user perceptions of AI tools can vary depending on the use scenario.
Yan et al. [67] explore differences in the user perceptions of ChatGPT across various sce-
narios, identifying factors such as information quality, perceived risk, attitude, and policy
support. They find that users’ perceptions significantly differ based on the context in
which they use AI tools, influencing their attitudes and support for policies regulating AI.
This suggests that, while AI has the potential to improve disaster management practices,
organizations must consider user acceptance and address concerns related to information
quality and risks.
Moreover, the integration of AI into organizational processes has implications for

workforce dynamics and organizational adaptation. Farrow [68] examine the implications
of different ratios of human and AI intelligence in future organizational operating models.
Their findings reveal concerns about potential job displacement and loss of human inter-
action, as well as opportunities for enhancing efficiency and problem-solving capabilities.
Organizations must balance the benefits of automation and efficiency with the need to
foster a culture of innovation and maintain employee engagement.
Additionally, ethical considerations play a crucial role in the successful implementation
of AI in organizational contexts. Cumming et al. [64] emphasize the importance of AI-
ethics-led sustainability frameworks, identifying key principles such as beneficence, non-
maleficence, justice, explainability, autonomy, privacy, and bias mitigation. They argue that
integrating ethical considerations into AI deployment is essential for building trust and
ensuring that AI contributes positively to sustainability goals. Our findings, which show
low prioritization in terms of those factors related to climate control and gender equality,
may reflect a gap in addressing ethical considerations within organizations.
In conclusion, the integration of AI, generative AI, and conversational agents into
scenario planning and disaster management holds significant potential for enhancing
organizational resilience and sustainability. Organizations should consider investing in
AI capabilities that augment human expertise, promote ethical practices, and foster a
culture of innovation. By doing so, they can leverage the strengths of both AI and human
intelligence to navigate uncertainties, enhance decision-making, and contribute to achieving
the Sustainable Development Goals.
5.7. Limitations and Future Research

Our study has some limitations. The correlations observed are based on self-reported
data from survey respondents, which may be subject to biases. Additionally, the sample
may not be representative of all organizations, limiting the generalizability of the findings.
Future research could explore the reasons behind the low prioritization of climate
control and gender equality, perhaps through qualitative methods such as interviews or
focus groups. Investigating the barriers to integrating these factors into organizational
strategies could inform the development of targeted interventions.
Moreover, further studies could examine the effectiveness of different risk manage-
ment practices in enhancing disaster resilience, particularly the role of AI and stakeholder
engagement. Longitudinal studies could assess the impact of these practices over time and
in different contexts.
6. Conclusions
Increasing an organization’s resilience to disasters through sustainable development
scenarios enables better adaptation to complex environmental conditions. The use of artifi-
cial intelligence may enrich and accelerate scenario creation, but ensuring professionalism
remains essential, including ethical considerations, critical evaluation, external expert
involvement, and thorough documentation.
Our research on Lithuanian companies revealed several key findings:
• Engagement in Sustainability and Disaster Prevention: Only 37% of the respondents
reported encountering sustainable development and disaster management issues, and
a similar proportion developed related scenarios. While external stakeholders (the
state, suppliers, and consumers) are often involved, 57% indicated that companies
ultimately address these challenges independently.
• Preparedness vs. Rapid Response: Companies prioritize rapid response and
liquidation of disaster consequences (4.5/5) over risk management and increas-
ing resilience (3.8/5). This suggests an emphasis on immediate action rather than
long-term preparedness.
• Psychological Stress and SDGs: The correlation and mutual information analyses un-
covered that employee psychological stress negatively correlates with the importance
assigned to SDGs during disasters. Thus, stress-reducing measures are crucial for
remaining focused on sustainability.
• Mutual Information Insights: The MI analysis complemented the correlation analy-
sis by detecting non-linear dependencies and identifying potential survey question
overlaps. For example, “Securing supply chains” (B3.5) emerged as a strong predictor.
The use of artificial intelligence in creating typical (pessimistic, most likely, and opti-
mistic) scenarios for sustainable development, including disaster impact assessments and
resilience strategies, could potentially enable resource-limited organizations to identify
priority areas for competence building. When performing scenario generation, it remains
important to consider ethics, logical consistency, contextual relevance, and to document
prompt engineering. Following these practices should increase the efficiency of ChatGPT
use in scenario generation. In this research, we provided detailed prompt engineering
recommendations and examples, demonstrating how to generate scenarios using GenAI.
The research might be useful for researchers and decision-makers. Although only
Lithuanian companies were studied, the insights developed from this research could be
used on a global scale considering the more diverse situations regarding organizations’ re-
silience to disasters and the probability and scale of the disasters themselves. In conclusion,
as the world grapples with an increasingly complex array of crises, this study provides
insights for integrating disaster risk management into sustainable development strategies.
The findings underscore the importance of building resilience at the organizational level
and reimagining “normality” in the face of continuous disruptions, thereby contributing
to the broader efforts to achieve the SDGs in a world where crises have become the new
normal [69].
Author Contributions: Conceptualization, A.M. and O.R.Š.; Data curation, A.M. and A.B.; Formal
analysis, A.B.; Investigation, O.R.Š.; Methodology, A.M., A.B. and O.R.Š.; Resources, A.M. and A.B.;
Software, A.B.; Supervision, O.R.Š.; Validation, A.M. and A.B.; Visualization, O.R.Š.; Writing—original
draft, A.M.; Writing—review and editing, A.B. All authors have read and agreed to the published
version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data were obtained through paper questionnaires.
Acknowledgments: The authors acknowledge the use of ChatGPT 4 (OpenAI, https://chat.openai.
com) to generate prompts that were provided in the article, to proofread the final draft, and to
improve language quality.
Conflicts of Interest: The authors declare no conflicts of interest.
Appendix A. The Conversation with ChatGPT

Here, we provide an exported conversation with ChatGPT [70] as an important part of
our investigation. Note that we use LaTeX to add structure to the text and remove possible
ambiguities as, from our observations, it is parsed by AI perfectly.
Prompt:
Using the detailed context and survey findings provided, generate three comprehensive
scenarios---pessimistic, most likely, and optimistic---that explore how a large manufacturing company in
Lithuania can enhance its disaster preparedness and sustainable development practices over the next
decade (2025 to 2035). Let's consider artificial company named ``Baltic Car Makers''.
Definition of a Good Scenario in This Context:
In the context of this research, a good scenario is a coherent, plausible, and detailed narrative that
describes a potential future situation involving specific challenges and outcomes related to sustainable
development and disaster management. Each scenario should:
\begin{itemize}
\item Incorporate Survey Findings and Contextual Data: Directly integrate the provided context
and survey results to ensure relevance and applicability.

\item Address Key Elements: Include analysis of possible disaster impacts and actions to
increase organizational resilience, considering stakeholders' actions.
\item Align with Sustainable Development Goals (SDGs): Reflect on all seventeen SDGs,
emphasizing those identified as important or less-prioritized in the survey.
\item Assess Potential Impacts and Probabilities: Evaluate the likelihood, scale, and extent of
various disasters and their potential effects on the company.
\item Detail Strategies for Resilience: Describe how the company can enhance its resilience to
disasters under each scenario, considering the specific challenges and opportunities presented.
\item Provide Actionable Insights: Offer practical recommendations for strategic planning,
disaster preparedness, and sustainable development.
\item Maintain Plausibility and Coherence: Ensure that the scenario is believable and logically
consistent, with a clear narrative flow.
\item Cover the Specified Timeframe: Encompass developments and projections from 2025 to 2035.
\end{itemize}
Context and Survey Findings:
\begin{itemize}
\item The majority of professionals in Lithuanian companies are engineers, designers, directors,
department heads, and managers, with most holding higher education qualifications.
\item Half of the companies are large enterprises, primarily in the manufacturing sector, with the
remainder being SMEs and micro-companies.
\item Only 37\% of companies have encountered issues related to sustainable development and disaster
management, and a similar percentage actively develop related scenarios.
\item Companies rarely initiate sustainable development and disaster prevention activities
independently; external stakeholders like the state, suppliers, and consumers are typically involved.
However, 57\% indicated that, in practice, only the company itself addresses these issues.
\item Large companies rate their disaster preparedness higher (3.5 out of 5) compared to SMEs and micro-
companies.
\item Differences exist in the prioritization of Sustainable Development Goals (SDGs) during disasters.
Clean water and sanitation, good health and well-being, zero hunger, life on land and below water, and
responsible consumption are deemed important across all company sizes. Micro-companies place less
emphasis on gender equality, climate action, and industry, innovation, and infrastructure.
\item The most critical scenarios for disaster management are rapid response and liquidation of disaster
consequences, securing supply chains, attracting and mobilizing stakeholders, and temporary
accommodation of residents.
\item Frequent negative impacts of disasters include electricity and communication disruptions, supplier
and customer issues, organizational challenges during pandemics, financial losses, human resources
problems, and psychological stress among employees.
\end{itemize}
Scenario Requirements:
Develop three scenarios---pessimistic, most likely, and optimistic---for the sustainable development of
a large manufacturing company in Lithuania. Each scenario should:
\begin{itemize}
\item Analyze the Impact of Possible Disasters:
\begin{itemize}
\item Assess the probability, size, and extent of potential disasters.
\item Evaluate how these disasters could affect the company's operations, supply chains, employees, and
stakeholders.
\end{itemize}
\item Detail Actions to Increase Organizational Resilience:
\begin{itemize}
\item Propose strategies and initiatives the company can implement to enhance its disaster preparedness
and sustainability practices.
\item Consider both internal efforts and collaboration with external stakeholders.
\end{itemize}
\item Consider Stakeholder Actions:
\begin{itemize}
\item Include potential actions and reactions from employees, local communities, suppliers, consumers,
government agencies, and other relevant parties.
\item Address how stakeholder engagement can support or hinder the company's efforts.
\end{itemize}
\item Integrate Sustainable Development Goals (SDGs):
\begin{itemize}
\item Reflect on all seventeen SDGs, highlighting specific actions to promote those identified as
important (e.g., clean water and sanitation, good health and well-being, responsible consumption).
\item Suggest initiatives to improve areas rated lower by micro-companies (e.g., gender equality,
climate action, industry innovation).
\end{itemize}
\item Describe Resilience Enhancements:
\begin{itemize}
\item Explain how the company can strengthen its resilience in each scenario, taking into account the
potential impacts of various disasters.
\item Include measures such as infrastructure upgrades, technology adoption, policy changes, and
workforce training.
\end{itemize}
\item Timeframe:
\begin{itemize}
\item Cover developments and projections from 2025 to 2035, outlining short-term and long-term
strategies.
\end{itemize}
\end{itemize}
Key Elements to Address in Each Scenario:
\begin{enumerate}
\item Company Profile:
\begin{itemize}
\item A large manufacturing company specializing in a specific industry (e.g., automotive parts,
electronics, machinery).
\item Emphasize the company's significance in the Lithuanian economy and the manufacturing sector.
\end{itemize}
\item Initial Challenges:
\begin{itemize}
\item Limited prior engagement in sustainable development and disaster management practices.
\item Increasing frequency of natural disasters due to climate change impacting operations.
\item Existing vulnerabilities such as electricity and communication disruptions, supplier and customer
issues, and employee stress.
\end{itemize}
\item Strategic Initiatives:
\begin{itemize}
\item Steps the company takes to develop and implement sustainable development and disaster response
strategies.
\item How the company leverages its internal resources despite initial reluctance to act independently.
\item Collaboration with external stakeholders to enhance preparedness and resilience.
\end{itemize}
\item Integration of Artificial Intelligence:
\begin{itemize}
\item Implementation of AI technologies, including machine learning and predictive analytics, in
disaster planning and management.
\item Use of AI for early warning systems, supply chain optimization, risk assessment, and real-time
decision-making during disasters.
\end{itemize}
\item Addressing Sustainable Development Goals:
\begin{itemize}
\item Specific actions to promote important SDGs.
\item Initiatives to improve areas rated lower by micro-companies.
\end{itemize}
\item Stakeholder Engagement:
\begin{itemize}
\item How the company engages employees, local communities, suppliers, consumers, and government
agencies.
\item Programs to reduce psychological stress among employees and enhance human resource management
during crises.
\end{itemize}
\item Overcoming Barriers:
\begin{itemize}
\item Strategies to address financial constraints and resource limitations.
\item Measures to mitigate frequent negative impacts of disasters identified in the survey.
\end{itemize}
\item Outcomes and Impact:
\begin{itemize}
\item Improvements in disaster preparedness ratings within the company.
\item Positive effects on operational continuity, profitability, and competitive advantage.
\item Contributions to the broader community's resilience and advancement toward sustainable development.
\end{itemize}
\item Future Outlook:
\begin{itemize}
\item Long-term benefits and lessons learned.
\item How the company's actions position it as a leader in sustainable practices and disaster management.
\end{itemize}
\end{enumerate}
Additional Guidance:
\begin{itemize}
\item Pessimistic Scenario:
\begin{itemize}
\item Explore a future where challenges intensify, and the company's efforts are met with significant
obstacles.
\item Consider worst-case outcomes, such as severe disasters, inadequate stakeholder support, and
limited success in implementing initiatives.
\end{itemize}
\item Most Likely Scenario:
\begin{itemize}
\item Present a realistic future based on current trends and moderate success in addressing challenges.
\item Include balanced outcomes with both achievements and setbacks.
\end{itemize}
\item Optimistic Scenario:

\begin{itemize}
\item Envision a future where the company successfully overcomes challenges and exceeds expectations.
\item Highlight best-case outcomes, such as effective disaster mitigation, strong stakeholder
collaboration, and significant advancements in sustainability.
\end{itemize}
\end{itemize}
Ensure that each scenario is
\begin{itemize}
\item Coherent and Plausible: The narrative should be logical and believable within the context
provided.
\item Detailed and Specific: Include concrete examples, data points, and projections.
\item Directly Incorporating Survey Findings: Use the provided context to ground the scenarios
in reality.
\item Actionable: Offer insights and recommendations that can inform strategic planning and
decision-making.
\end{itemize}
Provide answer using latex language.
Answer:
\section*{Pessimistic Scenario (2025--2035)}
\subsection*{Company Profile}
Baltic Car Makers is a large Lithuanian manufacturing company specializing in automotive parts
production. As a key player in the national economy, it employs a significant number of professionals,
including engineers, designers, and managers.
\subsection*{Initial Challenges}
Despite its size, the company has limited prior engagement in sustainable development and disaster
management practices. With climate change accelerating, natural disasters become more frequent, directly
impacting operations. Existing vulnerabilities include:
\begin{itemize} \item Frequent electricity and communication disruptions. \item Supply chain issues with
both suppliers and customers. \item Organizational challenges during pandemics. \item Psychological
stress among employees. \end{itemize}
\subsection*{Impact of Possible Disasters}
Between 2025 and 2035, Lithuania experiences an increase in severe natural disasters:
\begin{itemize} \item High Probability: Floods, storms, and heatwaves become more common due to
climate change. \item Size and Extent: Disasters affect large regions, disrupting national
infrastructure. \item Effects on the Company: \begin{itemize} \item Operations are frequently
halted. \item Supply chains are severely disrupted. \item Physical infrastructure suffers significant
damage. \item Employee well-being deteriorates due to constant crises. \end{itemize} \end{itemize}
\subsection*{Actions to Increase Organizational Resilience}
Baltic Car Makers makes minimal efforts to enhance disaster preparedness:
\begin{itemize} \item Internal Efforts: Reluctance to invest in sustainability due to financial

constraints. \item Lack of Initiative: Waits for external stakeholders to lead, but such
support is limited. \item Failure to Collaborate: Does not actively seek partnerships or
government assistance. \end{itemize}
\subsection*{Stakeholder Actions}
Stakeholder engagement is weak:
\begin{itemize} \item Employees: High stress levels lead to decreased productivity and
increased turnover. \item Suppliers and Customers: Face their own challenges, causing further
supply chain breakdowns. \item Government Agencies: Provide minimal support due to resource
limitations. \end{itemize}
\subsection*{Integration of Artificial Intelligence}
The company hesitates to invest in AI technologies:
\begin{itemize} \item Lack of Adoption: No implementation of AI for disaster planning or supply

chain management. \item Technological Lag: Competitors adopt AI, leaving Baltic Car Makers
behind. \end{itemize}
\subsection*{Addressing Sustainable Development Goals}
Minimal progress is made on SDGs:

\begin{itemize} \item Important SDGs Neglected: Little action on clean water (SDG 6), good
health (SDG 3), or responsible consumption (SDG 12). \item Less-Prioritized SDGs Ignored:
Gender equality (SDG 5), climate action (SDG 13), and innovation (SDG 9) remain unaddressed.
\end{itemize}
\subsection*{Resilience Enhancements}
The company fails to strengthen its resilience:
\begin{itemize} \item Infrastructure: No significant upgrades; facilities remain vulnerable.

\item Policies and Training: Lacks disaster preparedness policies; no employee training.
\end{itemize}
\subsection*{Overcoming Barriers}
Financial and operational barriers persist:
\begin{itemize} \item Financial Constraints: No efforts to secure funding or reallocate

resources. \item Negative Impacts Unmitigated: Disasters continue to cause significant
operational losses. \end{itemize}
\subsection*{Outcomes and Impact}
Negative outcomes prevail:
\begin{itemize} \item Disaster Preparedness Ratings: Remain low; company is ill-prepared for
crises. \item Operational Continuity: Frequent disruptions lead to declining productivity and
profitability. \item Competitive Advantage: Loses market share to better-prepared competitors.
\end{itemize}
\subsection*{Future Outlook}
The long-term outlook is bleak:
\begin{itemize} \item Survival: Company struggles to survive amidst mounting challenges. \item
Reputation: Damaged due to poor crisis management and sustainability neglect. \item
Missed Opportunities: Fails to capitalize on advancements in sustainable practices.
\end{itemize}
\newpage
\section*{Most Likely Scenario (2025--2035)}
Baltic Car Makers maintains its position as a leading automotive parts manufacturer in Lithuania,
contributing significantly to the economy and employment.
Recognizing the increasing frequency of natural disasters, the company begins to engage in sustainable
development and disaster management practices, albeit cautiously.
Natural disasters occur with moderate frequency:
\begin{itemize} \item Medium Probability: Occasional floods and storms impact operations. \item
Effects on the Company: \begin{itemize} \item Operations are temporarily disrupted. \item
Supply chains experience delays but not complete breakdowns. \item Employee well-being is affected but
managed. \end{itemize} \end{itemize}
The company takes moderate steps:
\begin{itemize} \item Internal Efforts: \begin{itemize} \item Forms a basic disaster response
team. \item Allocates limited funds for infrastructure improvements. \end{itemize} \item
Collaboration: \begin{itemize} \item Engages with suppliers to strengthen supply chains. \item
Participates in industry forums on disaster preparedness. \end{itemize} \end{itemize}
Stakeholder engagement improves:
\begin{itemize} \item Employees: Receive some training; morale stabilizes. \item

Suppliers and Customers: Work together to mitigate disruptions. \item \textbf{Government
Agencies}: Provide occasional support, though bureaucratic hurdles persist. \end{itemize}

Partial adoption of AI technologies:
\begin{itemize} \item Implementation: \begin{itemize} \item Introduces AI for basic supply

chain management. \item Uses predictive analytics for maintenance scheduling. \end{itemize} \item
Benefits: Some operational efficiencies are realized. \end{itemize}
Moderate progress is made:
\begin{itemize} \item Important SDGs: \begin{itemize} \item Implements water-saving measures

(SDG 6). \item Enhances workplace safety (SDG 3). \end{itemize} \item Less-Prioritized SDGs:
\begin{itemize} \item Begins to explore climate action initiatives (SDG 13). \item Invests modestly in
innovation (SDG 9). \end{itemize} \end{itemize}
Strengthens resilience to a moderate extent:
\begin{itemize} \item Infrastructure: Upgrades critical systems to withstand common disasters.

\item Policies and Training: Implements basic disaster preparedness policies. \end{itemize}
Some barriers are addressed:
\begin{itemize} \item Financial Constraints: Secures modest funding through partnerships. \item
Negative Impacts: Reduces operational disruptions through improved planning. \end{itemize}
Mixed outcomes are observed:
\begin{itemize} \item Disaster Preparedness Ratings: Improve to an average level. \item

Operational Continuity: Stability is achieved, though not optimal. \item \textbf{Competitive
Advantage}: Maintains market position but does not significantly excel. \end{itemize}
A cautious optimism prevails:
\begin{itemize} \item Sustainability: Recognizes the need for ongoing efforts. \item
Leadership Position: Seen as a responsible company, though not a leader. \item \textbf{Lessons
Learned}: Understands the importance of proactive measures. \end{itemize}
\newpage
\section*{Optimistic Scenario (2025--2035)}
Baltic Car Makers emerges as a leader in sustainable manufacturing and disaster preparedness, setting
industry benchmarks in Lithuania and beyond.
Proactively addresses sustainable development and disaster management, transforming challenges into
opportunities.
Natural disasters have minimal impact due to effective strategies:
\begin{itemize} \item High Probability: Disasters occur but are mitigated effectively. \item
Effects on the Company: \begin{itemize} \item Operations remain uninterrupted. \item Supply
chains are resilient and adaptive. \item Employee well-being is prioritized and maintained.
\end{itemize} \end{itemize}
Significant initiatives are undertaken:
\begin{itemize} \item Internal Efforts: \begin{itemize} \item Establishes a Sustainability and

Resilience Department. \item Integrates SDGs into corporate strategy. \end{itemize} \item
Investment: \begin{itemize} \item Allocates 10% of annual revenue to sustainability
initiatives. \item Invests in renewable energy, reducing dependence on external grids. \end{itemize}
\item Collaboration: \begin{itemize} \item Partners with universities for research on
sustainable materials. \item Engages with international organizations for best practices. \end{itemize}
\end{itemize}
Strong engagement leads to collective success:
\begin{itemize} \item Employees: Highly engaged; benefit from wellness programs and training.
\item Suppliers and Customers: Integrated into disaster preparedness plans. \item
Government Agencies: Provide support and recognize the company's leadership. \end{itemize}
Advanced AI technologies are fully implemented:
\begin{itemize} \item Implementation: \begin{itemize} \item Uses AI for predictive analytics

and early warning systems. \item Optimizes supply chains and real-time decision-making during disasters.
\end{itemize} \item Benefits: Achieves high efficiency and swift response to incidents.
\end{itemize}
Significant progress across all SDGs:
\begin{itemize} \item SDG 6 (Clean Water): Implements water recycling in manufacturing. \item
SDG 3 (Good Health): Offers comprehensive health programs. \item \textbf{SDG 12 (Responsible
Consumption)}: Adopts circular economy principles. \item SDG 5 (Gender Equality): Increases
female representation in leadership. \item SDG 13 (Climate Action): Achieves carbon neutrality
by 2030. \item SDG 9 (Industry, Innovation): Invests in cutting-edge technologies. \end{itemize}
Company's resilience is greatly enhanced:
\begin{itemize} \item Infrastructure: Builds disaster-resistant facilities with renewable

energy sources. \item Policies and Training: Establishes comprehensive policies; employees are
well-prepared. \end{itemize}
Effectively overcomes all barriers:
\begin{itemize} \item Financial Constraints: \begin{itemize} \item Secures funding through

green bonds and EU grants. \item Leverages investments and partnerships. \end{itemize} \item
Negative Impacts: \begin{itemize} \item Implements robust cybersecurity measures. \item
Develops alternative supply chain strategies. \end{itemize} \end{itemize}
Positive outcomes are achieved:
\begin{itemize} \item Disaster Preparedness Ratings: Ranks in the top 5% in Europe. \item
Operational Continuity: Maintains uninterrupted operations. \item \textbf{Competitive
Advantage}: Increases market share by 25%. \item Community Impact: Supports local communities
through sustainable projects. \end{itemize}
The company is well-positioned for the future:
\begin{itemize} \item Leadership: Receives international recognition for sustainability. \item

Innovation: Continues to innovate with sustainable materials. \item Growth: Expands
into new markets and sectors. \end{itemize}
Prompt:
Could you comment where and how the data of our study findings was reflected in your scenarios?
Answer:
1. Professional Roles and Company Sizes
Study Finding: The majority of professionals in Lithuanian companies are engineers, designers,
directors, department heads, and managers, with most holding higher education qualifications. Half of
the companies are large enterprises, primarily in the manufacturing sector.
Reflection in Scenarios:
In all three scenarios, Company Profile sections emphasize that Baltic Car Makers is a large
manufacturing company specializing in automotive parts, highlighting its significance in the Lithuanian
economy.
The company is described as employing a significant number of professionals, including engineers,
designers, and managers, mirroring the professional roles indicated in the study.
2. Limited Prior Engagement in Sustainable Development and Disaster Management

Study Finding: Only 37% of companies have encountered issues related to sustainable development and
disaster management, and a similar percentage actively develop related scenarios.
Pessimistic Scenario: The company has limited prior engagement in sustainable development and disaster
management practices, directly reflecting the low percentage from the study.
Most Likely Scenario: The company begins to engage in these practices cautiously, indicating moderate
involvement.
Optimistic Scenario: The company proactively addresses sustainable development and disaster management,
exceeding the current engagement levels indicated in the study.
3. Initiation of Activities and Reliance on External Stakeholders
Study Finding: Companies rarely initiate sustainable development and disaster prevention activities
independently; external stakeholders are typically involved. However, 57% indicated that, in practice,
only the company itself addresses these issues.
Pessimistic Scenario: The company fails to initiate sustainability efforts independently and waits for
external stakeholders, reflecting the reliance indicated in the study.
Most Likely Scenario: The company takes moderate steps and collaborates with external stakeholders,
balancing internal and external efforts.
Optimistic Scenario: The company leads initiatives independently while also engaging stakeholders,
embodying the proactive approach of the 57% who indicated that companies address issues themselves.
4. Disaster Preparedness Ratings
Study Finding: Large companies rate their disaster preparedness higher (3.5 out of 5) compared to SMEs
and micro-companies.
Pessimistic Scenario: The company's preparedness remains low due to minimal efforts, indicating a rating
below the average.
Most Likely Scenario: The company's preparedness ratings improve to an average level, aligning with the
3.5 out of 5 rating for large companies.
Optimistic Scenario: The company achieves top ratings, surpassing the average and becoming a leader in
disaster preparedness.
5. Prioritization of Sustainable Development Goals (SDGs)
Study Finding: Important SDGs include clean water and sanitation (SDG 6), good health and well-being
(SDG 3), and responsible consumption (SDG 12). Micro-companies place less emphasis on gender equality
(SDG 5), climate action (SDG 13), and industry, innovation, and infrastructure (SDG 9).
Pessimistic Scenario: Minimal progress on important SDGs; important SDGs are neglected, and less
prioritized SDGs are ignored.
Most Likely Scenario: Moderate progress on important SDGs, with initial efforts to address less
prioritized ones.
Optimistic Scenario: Significant progress across all SDGs, including those less emphasized,
demonstrating comprehensive commitment.
6. Impact of Possible Disasters
Study Finding: Frequent negative impacts include electricity and communication disruptions, supplier and
customer issues, organizational challenges during pandemics, financial losses, human resources problems,
and psychological stress among employees.
Pessimistic Scenario: Disasters have severe impacts due to lack of preparedness, mirroring the frequent
negative impacts identified.
Most Likely Scenario: Disasters cause temporary disruptions, but impacts are managed to some extent.
Optimistic Scenario: Disasters have minimal impact due to effective strategies, reflecting the potential
to overcome the negative impacts highlighted in the study.
7. Actions to Increase Organizational Resilience
Study Finding: Large companies rate their preparedness higher and may have more resources to enhance
resilience.
Pessimistic Scenario: The company makes minimal efforts to enhance resilience, failing to leverage its
resources.
Most Likely Scenario: The company takes moderate steps to improve resilience through basic disaster
response teams and infrastructure improvements.
Optimistic Scenario: The company undertakes significant initiatives, such as establishing a

Sustainability and Resilience Department and investing heavily in infrastructure and technology, fully
leveraging its capacity as a large enterprise.
8. Stakeholder Actions and Engagement
Study Finding: External stakeholders like the state, suppliers, and consumers are typically involved,
but companies may also need to act independently.
Pessimistic Scenario: Weak stakeholder engagement, with little collaboration or support.

Most Likely Scenario: Improved engagement with suppliers and participation in industry forums.
Optimistic Scenario: Strong engagement with all stakeholders, including employees, local communities,
suppliers, consumers, and government agencies, enhancing collective resilience.
9. Integration of Artificial Intelligence
Study Finding: While not explicitly mentioned in the survey, the integration of AI is a key element in
modern disaster preparedness and aligns with SDG 9 (Industry, Innovation, and Infrastructure).
Pessimistic Scenario: Hesitation to invest in AI technologies, leading to a technological lag.

Most Likely Scenario: Partial adoption of AI for basic supply chain management and predictive analytics.
Optimistic Scenario: Full implementation of advanced AI technologies for disaster planning, early
warning systems, supply chain optimization, and real-time decision-making, showcasing innovation and
infrastructure development.
10. Overcoming Barriers
Study Finding: Financial constraints and resource limitations are significant barriers to implementing
sustainable practices and enhancing disaster preparedness.
Pessimistic Scenario: The company makes no efforts to overcome financial constraints; barriers persist.
Most Likely Scenario: The company secures modest funding through partnerships and addresses some
barriers.
Optimistic Scenario: The company effectively overcomes barriers by securing funding through green bonds
and EU grants, and by leveraging investments and partnerships.
11. Addressing Employee Well-being and Human Resource Challenges
Study Finding: Psychological stress among employees and human resources problems are frequent negative
impacts of disasters.
Pessimistic Scenario: Employee stress increases due to constant crises, with decreased productivity and
high turnover.
Most Likely Scenario: The company provides some training and stabilizes morale, partially addressing
employee well-being.
Optimistic Scenario: The company offers comprehensive health and wellness programs, leading to high
employee engagement and well-being.
12. Securing Supply Chains
Study Finding: Securing supply chains is one of the most critical scenarios for disaster management.
Pessimistic Scenario: Supply chains are severely disrupted, causing operational breakdowns.
Most Likely Scenario: Supply chains experience delays but not complete breakdowns, with the company
working with suppliers to mitigate issues.
Optimistic Scenario: Supply chains are resilient and adaptive, maintained through strategic partnerships
and AI optimization.
13. Rapid Response and Liquidation of Disaster Consequences
Study Finding: Rapid response is critical in disaster management scenarios.
Pessimistic Scenario: The company fails to respond effectively, resulting in prolonged disruptions.
Most Likely Scenario: The company implements basic disaster response measures, achieving temporary
stability.
Optimistic Scenario: The company excels in rapid disaster response, minimizing impacts and quickly
returning to normal operations.
14. Addressing Less-Prioritized SDGs by Micro-Companies

Study Finding: Micro-companies place less emphasis on gender equality, climate action, and industry,
innovation, and infrastructure.
Pessimistic Scenario: The company ignores these less-prioritized SDGs, similar to micro-companies.
Most Likely Scenario: The company begins to explore initiatives in these areas, indicating growing
awareness.
Optimistic Scenario: The company actively promotes gender equality (e.g., increasing female leadership),
commits to climate action (e.g., achieving carbon neutrality), and invests in innovation and
infrastructure, addressing the gaps identified in the study.
15. Temporary Accommodation of Residents
Study Finding: Temporary accommodation of residents is a critical scenario in disaster management.
Pessimistic Scenario: The company does not participate in community support efforts during disasters.
Most Likely Scenario: The company provides limited support to local communities.
Optimistic Scenario: The company actively supports local communities by offering resources and
facilities for temporary accommodation, enhancing community resilience.
Appendix B. The Full Heatmap of Correlations

Here, we provide correlation coefficients between all the scores in the form of a
heatmap for the reader.
C1.13 - Injuries to employees 1.0
C1.7 - Fire in the company
C1.19 - Financial losses
C1.6 - Destruction in the company
C1.14 - Deaths of employees
C1.10 - Problems with work organization due to the pandemic
C1.9 - Absence of heating and ventilation 0.8
C1.15 - Information system violations due to hacking and viruses
C1.3 - Interruption of food supply
C1.5 - Disconnection of communication
C1.2 - Water supply interruption
C1.8 - A flood in the company
C1.11 - Problems with suppliers 0.6
C1.1 - Disconnection of electricity
C1.16 - Psychological stress of employees
C1.20 - Human resources issues
C1.17 - Legal disputes
C1.18 - Damage to reputation
C1.4 - Closed roads 0.4
C1.12 - Problems with customers
B3.4 - Attracting and mobilizing stakeholders
B3.1 - Risk management and increasing of disaster resilience
B3.8 - Data storage and protection
B2.12 - Responsible consumption & production
B2.13 - Climate action 0.2
B1.1 - Creativity of company employees
B1.2 - Entrepreneurship of company employees
B3.3 - The use of artificial intelligence in disaster planning and management
B2.4 - Quality education
B3.6 - Ensuring sustainability in the event of disasters
B2.6 - Clean water and sanitation 0.0
B2.3 - Good health and well-being
B3.2 - Rapid response and liquidation of disaster consequences
B2.5 - Gender equality
B2.8 - Decent work and economic growth
B2.9 - Industry, innovation, and infrastructure 0.2
B2.15 - Life on land
B3.7 - Temporary accommodation of residents
B2.7 - Affordable and clean energy
B2.11 - Sustainable cities and communities
B2.2 - Zero hunger 0.4
B2.1 - No poverty
B2.16 - Peace, justice & strong institutions
B2.17 - Partnerships for the goals
C1.1
C1.7
C1.1
C1.6
C1.1
C1.1
C1.9
C1.1
C1.3
C1.5
C1.2
C1.8
C1.1
C1.1
C1.1
C1.2
C1.1
C1.1
C1.4
B3.1
B3.4
B3.1
B3.5
B2.8
B2.1
B1.1
B1.1
B3.2
B2.3
B3.4
B2.6
B2.6
B3.3
B2.2
B2.1
B2.5
B2.8
B2.9
B2.1
B3.1
B2.7
B2.7
B2.1
B2.2
B2.1
B2.1
.1
C1
Figure A1. The full heatmap of correlations between scores for different questions.
Appendix C. The Main Statistics Parameters for Responses

Here, we provide the main statistics parameters for responses: mean and standard
deviation (SD).
Table A1. The estimation of the creativity and entrepreneurship of company employees, average and
standard deviation values.
Parameter Mean SD
B1.1—Creativity of company employees 3.533 1.196
B1.2—Entrepreneurship of company employees 3.400 1.003
Table A2. The importance of different SDGs for sustainable development in the context of disaster,
average and standard deviation values.
Parameter Mean SD
B2.1—No poverty 3.433 1.135
B2.2—Zero hunger 4.367 0.999
B2.3—Good health and well-being 4.400 0.855
B2.4—Quality education 3.833 1.117
B2.5—Gender equality 2.800 1.349
B2.6—Clean water and sanitation 4.733 0.521
B2.7—Affordable and clean energy 3.700 1.368
B2.8—Decent work and economic growth 4.133 1.196
B2.9—Industry, innovation, and infrastructure 4.033 1.217
B2.10—Reduced inequalities 3.300 1.343
B2.11—Sustainable cities and communities 3.733 1.258
B2.12—Responsible consumption and production 4.167 1.206
B2.13—Climate action 3.500 1.503
B2.14—Life below water 4.300 1.088
B2.15—Life on land 4.367 1.066
B2.16—Peace, justice and strong institutions 4.167 0.986
B2.17—Partnerships for the goals 4.167 1.341
Table A3. The estimation of the relevance of possible scenario types for managing catastrophes,
average and standard deviation values.
Parameter Mean SD
B3.1—Risk management and increasing disaster resilience 3.767 1.305
B3.2—Rapid response and liquidation of disaster consequences 4.500 1.042
B3.3—The use of artificial intelligence in disaster
3.133 1.196
planning and management
B3.4—Attracting and mobilizing stakeholders 4.033 1.217
B3.5—Securing supply chains 4.100 1.269
B3.6—Ensuring sustainability in the event of disasters 3.200 1.186
B3.7—Temporary accommodation of residents 4.133 1.167
B3.8—Data storage and protection 4.000 1.050
Table A4. The frequency of negative impacts during catastrophes in companies, average and standard
deviation values.
Parameter Mean SD
C1.1—Disconnection of electricity 2.867 1.252
C1.2—Water supply interruption 1.933 1.048
C1.3—Interruption of food supply 1.300 0.651
C1.4—Closed roads 2.033 1.326
C1.5—Disconnection of communication 2.433 1.135
Table A4. Cont.
Parameter Mean SD
C1.6—Destruction in the company 1.300 0.837
C1.7—Fire in the company 1.533 1.224
C1.8—A flood in the company 1.067 0.365
C1.9—Absence of heating and ventilation 2.100 1.062
C1.10—Problems with work organization due to the pandemic 2.300 1.264
C1.11—Problems with suppliers 2.733 1.172
C1.12—Problems with customers 2.700 1.368
C1.13—Injuries to employees 2.167 1.416
C1.14—Deaths of employees 1.433 1.165
C1.15—Information system violations due to hacking and viruses 2.033 1.159
C1.16—Psychological stress of employees 2.633 1.245
C1.17—Legal disputes 1.633 1.098
C1.18—Damage to reputation 1.600 1.102
C1.19—Financial losses 2.300 1.149
C1.20—Human resource issues 2.300 1.368
References
1. Bjerde, A.; Mountfield, E. Unlocking New Crisis Response Tools to Build a More Resilient Future. Available online: https://blogs.
worldbank.org/en/voices/unlocking-new-crisis-response-tools-build-more-resilient-future/ (accessed on 1 February 2024).
2. United Nations. THE 17 GOALS. Available online: https://sdgs.un.org/goals (accessed on 1 February 2024).
3. Cvetković, V.M.; Renner, R.; Aleksova, B.; Lukić, T. Geospatial and Temporal Patterns of Natural and Man-Made (Technological)
Disasters (1900–2024): Insights from Different Socio-Economic and Demographic Perspectives. Appl. Sci. 2024, 14, 8129.
[CrossRef]
4. Kulba, V.; Bakhtadze, N.; Zaikin, O.; Shelkov, A.; Chernov, I. Scenario analysis of management processes in the prevention and
the elimination of consequences of man-made disasters. Procedia Comput. Sci. 2017, 112, 2066–2075. [CrossRef]
5. Riddell, G.A.; van Delden, H.; Maier, H.R.; Zecchin, A.C. Exploratory scenario analysis for disaster risk reduction: Considering
alternative pathways in disaster risk assessment. Int. J. Disaster Risk Reduct. 2019, 39, 101230. . [CrossRef]
6. Sultan, S.M.N.; Maharjan, K.L. Cyclone-Induced Disaster Loss Reduction by Social Media: A Case Study on Cyclone Amphan in
Koyra Upazila, Khulna District, Bangladesh. Sustainability 2022, 14, 3909. [CrossRef]
7. de Goër de Herve, M. Near or distant time horizons? The determinants of the integration of long-term perspectives in disaster
risk management evaluation. Prog. Disaster Sci. 2024, 24, 100365. [CrossRef]
8. Urbanelli, A.; Frisiello, A.; Bruno, L.; Rossi, C. The ERMES chatbot: A conversational communication tool for improved emergency
management and disaster risk reduction. Int. J. Disaster Risk Reduct. 2024, 112, 104792. [CrossRef]
9. Vecchiato, R. Scenario planning, cognition, and strategic investment decisions in a turbulent environment. Long Range Plan. 2019,
52, 101865. [CrossRef]
10. Zalik, A. Oil ‘futures’: Shell’s Scenarios and the social constitution of the global oil market. Geoforum 2010, 41, 553–564. [CrossRef]
11. Jefferson, M. Shell scenarios: What really happened in the 1970s and what may be learned for current world prospects. Technol.
Forecast. Soc. Chang. 2012, 79, 186–197. [CrossRef]
12. Jefferson, M. Scenario planning: Evidence to counter ‘Black box’claims. Technol. Forecast. Soc. Chang. 2020, 158, 120156. [CrossRef]
13. Jefferson, M. A History of Energy and Societal Scenarios for a World in Transition: Fifty Years of Personal Experience with Shell
and Other Organisations. Energy Res. Soc. Sci. 2022, 90, 102609. [CrossRef]
14. Nguyen, D.N.; Usuda, Y.; Imamura, F. Gaps in and Opportunities for Disaster Risk Reduction in Urban Areas Through
International Standardization of Smart Community Infrastructure. Sustainability 2024, 16, 9586. [CrossRef]
15. Ameli, M.; Esfandabadi, Z.S.; Sadeghi, S.; Ranjbari, M.; Zanetti, M.C. COVID-19 and Sustainable Development Goals (SDGs):
Scenario analysis through fuzzy cognitive map modeling. Gondwana Res. 2023, 114, 138–155. [CrossRef] [PubMed]
16. Cernev, T. Global sustainability targets: Planetary boundary, global catastrophic risk, and disaster risk reduction considerations.
Prog. Disaster Sci. 2022, 16, 100264. [CrossRef]
17. Fang, C.; Cui, X.; Li, G.; Bao, C.; Wang, Z.; Ma, H.; Sun, S.; Liu, H.; Luo, K.; Ren, Y. Modeling regional sustainable development
scenarios using the Urbanization and Eco-environment Coupler: Case study of Beijing-Tianjin-Hebei urban agglomeration, China.
Sci. Total. Environ. 2019, 689, 820–830. [CrossRef] [PubMed]
18. Aguiar, A.P.D.; Collste, D.; Harmáčková, Z.V.; Pereira, L.; Selomane, O.; Galafassi, D.; Van Vuuren, D.; Van Der Leeuw, S.
Co-designing global target-seeking scenarios: A cross-scale participatory process for capturing multiple perspectives on pathways
to sustainability. Glob. Environ. Chang. 2020, 65, 102198. [CrossRef]
19. Wilts, R.; Britz, W. Distribution matters: Long-term quantification of the Sustainable Development Goals with household detail
for different socio-economic pathways. Glob. Transit. 2024, 6, 173–186. [CrossRef]
20. Allen, C.; Metternicht, G.; Wiedmann, T. National pathways to the Sustainable Development Goals (SDGs): A comparative
review of scenario modelling tools. Environ. Sci. Policy 2016, 66, 199–207. [CrossRef]
21. Beynaghi, A.; Trencher, G.; Moztarzadeh, F.; Mozafari, M.; Maknoon, R.; Leal Filho, W. Future sustainability scenarios for
universities: Moving beyond the United Nations Decade of Education for Sustainable Development. J. Clean. Prod. 2016,
112, 3464–3478. [CrossRef]
22. Nieto-Romero, M.; Milcu, A.; Leventon, J.; Mikulcak, F.; Fischer, J. The role of scenarios in fostering collective action for sustainable
development: Lessons from central Romania. Land Use Policy 2016, 50, 156–168. [CrossRef]
23. Kiviluoto, K.; Tapio, P.; Tuominen, A.; Lyytimäki, J.; Ahokas, I.; Silonsaari, J.; Schwanen, T. Towards sustainable mobility—
Transformative scenarios for 2034. Transp. Res. Interdiscip. Perspect. 2022, 16, 100690. [CrossRef]
24. Dahmen, J.; Kayaalp, M.E.; Ollivier, M.; Pareek, A.; Hirschmann, M.T.; Karlsson, J.; Winkler, P.W. Artificial intelligence bot
ChatGPT in medical research: The potential game changer as a double-edged sword. Knee Surg. Sport. Traumatol. Arthrosc. 2023,
31, 1187–1189. [CrossRef]
25. Petrudi, S.H.H.; Ahmadi, H.B.; Rehman, A.; Liou, J.J. Assessing suppliers considering social sustainability innovation factors
during COVID-19 disaster. Sustain. Prod. Consum. 2021, 27, 1869–1881. [CrossRef]
26. Montero, J.M.; Naimy, V.; Abi Farraj, N.; El Khoury, R. Natural disasters, stock price volatility in the property-liability insurance
market and sustainability: An unexplored link. Socio-Econ. Plan. Sci. 2024, 91, 101791. [CrossRef]
27. Johri, A.; Singh, R.K.; Alhumoudi, H.; Alakkas, A. Examining the Influence of Sustainable Management Accounting on Sustainable
Corporate Governance: Empirical Evidence. Sustainability 2024, 16, 9605. [CrossRef]
28. Kong, J.; Zhang, C.; Simonovic, S.P. Resilience and risk-based restoration strategies for critical infrastructure under uncertain
disaster scenarios. Sustain. Cities Soc. 2023, 92, 104510. [CrossRef]
29. Tang, J.; Wu, S.; Yang, S.; Shi, Y. Resilience Assessment of Urban Road Transportation in Rainfall. Remote Sens. 2024, 16, 3311.
[CrossRef]
30. Hao, H.; Wang, Y.; Chen, J. Empowering Scenario Planning with Artificial Intelligence: A Perspective on Building Smart and
Resilient Cities. Engineering 2024, in press. [CrossRef]
31. Cheng, C.S.; Behzadan, A.; Noshadravan, A. A post-hurricane building debris estimation workflow enabled by uncertainty-aware
AI and crowdsourcing. Int. J. Disaster Risk Reduct. 2024, 112, 104785. [CrossRef]
32. Hostetter, H.; Naser, M.; Huang, X.; Gales, J. The role of large language models (AI chatbots) in fire engineering: An examination
of technical questions against domain knowledge. Nat. Hazards Res. 2024, 4, 669–688. [CrossRef]
33. Maceika, A.; Bugajev, A.; Šostak, O.R. Evaluating Modular House Construction Projects: A Delphi Method Enhanced by
Conversational AI. Buildings 2024, 14, 1696. [CrossRef]
34. Spaniol, M.J.; Rowland, N.J. AI-assisted scenario generation for strategic planning. Futur. Foresight Sci. 2023, 5, e148. [CrossRef]
35. Rodgers, J.L.; Nicewander, W.A. Thirteen Ways to Look at the Correlation Coefficient. Am. Stat. 1988, 42, 59–66. [CrossRef]
36. Cover, T.M. Elements of Information Theory; John Wiley & Sons: Hoboken, NJ, USA, 1999.
37. Ward, J.H., Jr. Hierarchical Grouping to Optimize an Objective Function. J. Am. Stat. Assoc. 1963, 58, 236–244. [CrossRef]
38. Fahmi, F.Z.; Krismiyaningsih, E.; Sagala, S.A.H.; Rustiadi, S. Creative industries and disaster resilience: A focus on arts- and
culture-based industries in Indonesia. Int. J. Disaster Risk Reduct. 2023, 99, 104136. [CrossRef]
39. van Vuuren, D.P.; Zimm, C.; Busch, S.; Kriegler, E.; Leininger, J.; Messner, D.; Nakicenovic, N.; Rockstrom, J.; Riahi, K.; Sperling,
F.; et al. Defining a sustainable development target space for 2030 and 2050. One Earth 2022, 5, 142–156. [CrossRef]
40. Dai, D.; Bo, M.; Ren, X.; Dai, K. Application and exploration of artificial intelligence technology in urban ecosystem-based disaster
risk reduction: A scoping review. Ecol. Indic. 2024, 158, 111565. [CrossRef]
41. Ghaffarian, S.; Taghikhah, F.R.; Maier, H.R. Explainable artificial intelligence in disaster risk management: Achievements and
prospective futures. Int. J. Disaster Risk Reduct. 2023, 98, 104123. [CrossRef]
42. Maiti, M.; Kayal, P. Exploring innovative techniques for damage control during natural disasters. J. Saf. Sci. Resil. 2024, 5, 147–155.
[CrossRef]
43. Albahri, A.; Khaleel, Y.L.; Habeeb, M.A.; Ismael, R.D.; Hameed, Q.A.; Deveci, M.; Homod, R.Z.; Albahri, O.; Alamoodi, A.;
Alzubaidi, L. A systematic review of trustworthy artificial intelligence applications in natural disasters. Comput. Electr. Eng. 2024,
118, 109409. [CrossRef]
44. Aljadani, A. Assessing financial risk with extreme value theory: US financial indemnity loss data analysis. Alex. Eng. J. 2024,
105, 496–507. [CrossRef]
45. Puri, A.; Elkharboutly, M.; Ali, N.A. Identifying major challenges in managing post-disaster reconstruction projects: A critical
analysis. Int. J. Disaster Risk Reduct. 2024, 107, 104491. [CrossRef]
46. Hosseini, S.A.; Farahzadi, L.; Pons, O. Assessing the sustainability index of different post-disaster temporary housing unit
configuration types. J. Build. Eng. 2021, 42, 102806. [CrossRef]
47. Kourtit, K.; Nijkamp, P.; Banica, A. An analysis of natural disasters’ effects–A global comparative study of ‘Blessing in Disguise’.
Socio-Econ. Plan. Sci. 2023, 88, 101599. [CrossRef]
48. Otsuyama, K. How does the housing recovery and relocation process disconnect or reconnect neighborhood associations in
disaster public housing? A mixed-method study of a tsunami-impacted city in Japan. Int. J. Disaster Risk Reduct. 2023, 98, 104091.
[CrossRef]
49. Liu, H.; Tatano, H.; Kajitani, Y.; Yang, Y. Analysis of the influencing factors on industrial resilience to flood disasters using a
semi-markov recovery model: A case study of the Heavy Rain Event of July 2018 in Japan. Int. J. Disaster Risk Reduct. 2022,
82, 103384. [CrossRef]
50. Bakhshi Sasi, M.; Sarker, R.A.; Essam, D.L. A novel heuristic algorithm for disruption mitigation in a global food supply chain.
Comput. Ind. Eng. 2024, 194, 110334. [CrossRef]
51. Noaman, D.S.; El-Ghafour, S.A. Holistic design of energy-efficient temporary houses: Meeting ventilation, heating, cooling, and
lighting demands. J. Build. Eng. 2024, 86, 108534. [CrossRef]
52. Shin, W.; Park, J.; Kim, H.W. Online consumer demand during and after natural disasters: Evidence from the COVID-19 pandemic.
Inf. Manag. 2024, 61, 104006. [CrossRef]
53. Arroyabe, M.F.; Arranz, C.F.; Fernandez De Arroyabe, I.; Fernandez de Arroyabe, J.C. Exploring the economic role of cybersecurity
in SMEs: A case study of the UK. Technol. Soc. 2024, 78, 102670. [CrossRef]
54. Assis, T.; Cordeiro, F.; Schiavon, L. How stock market reacts to environmental disasters and judicial decisions: A case study of
Mariana’s dam collapse in Brazil. Int. Rev. Law Econ. 2023, 73, 106105. [CrossRef]
55. Šontaitė-Petkevičienė, M. Crisis management to avoid damage for corporate reputation: The case of retail chain crisis in the
Baltic countries. Procedia-Soc. Behav. Sci. 2014, 156, 452–457. [CrossRef]
56. Reynolds, L.; McDonell, K. Prompt programming for large language models: Beyond the few-shot paradigm. In Proceedings of
the Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing Systems, Yokohama, Japan, 8–13 May 2021;
pp. 1–7.
57. Jobin, A.; Ienca, M.; Vayena, E. The global landscape of AI ethics guidelines. Nat. Mach. Intell. 2019, 1, 389–399. [CrossRef]
58. Brown, T.B. Language models are few-shot learners. arXiv 2020, arXiv:2005.14165.
59. Wulf, T.; Meißner, P.; Stubner, S. A scenario-based approach to strategic planning: Integrating planning and process perspective of strategy;
HHL Leipzig Graduate School of Management: Leipzig, Germany, 2010.
60. Sætra, H.S. Science fiction, sustainability, and scenario use: Comprehensive scenarios for improved strategy development and
innovation. Technovation 2024, 132, 102976. [CrossRef]
61. Heo, S.; Byun, J.; Ifaei, P.; Ko, J.; Ha, B.; Hwangbo, S.; Yoo, C. Towards mega-scale decarbonized industrial park (Mega-
DIP): Generative AI-driven techno-economic and environmental assessment of renewable and sustainable energy utilization in
petrochemical industry. Renew. Sustain. Energy Rev. 2024, 189, 113933. [CrossRef]
62. Rohde, F.; Wagner, J.; Meyer, A.; Reinhard, P.; Voss, M.; Petschow, U.; Mollen, A. Broadening the perspective for sustainable
artificial intelligence: Sustainability criteria and indicators for Artificial Intelligence systems. Curr. Opin. Environ. Sustain. 2024,
66, 101411. [CrossRef]
63. Bag, S.; Sabbir Rahman, M.; Rogers, H.; Srivastava, G.; Harm Christiaan Pretorius, J. Climate change adaptation and disaster risk
reduction in the garment industry supply chain network. Transp. Res. Part E Logist. Transp. Rev. 2023, 171, 103031. [CrossRef]
64. Cumming, D.; Saurabh, K.; Rani, N.; Upadhyay, P. Towards AI ethics-led sustainability frameworks and toolkits: Review and
research agenda. J. Sustain. Financ. Account. 2024, 1, 100003. [CrossRef]
65. Kim, M.K.; Seo, J.; Kim, N.; Chung, J.B. Emerging industrial clusters of disaster safety industry in Korea. Heliyon 2023, 9, e17939.
[CrossRef]
66. Jiang, X.; Wang, X.; Fang, Y.; Yang, L.; Tatano, H. An agent-based model for the assessment of the ripple effect of disaster economic
losses considering firms’ adaptive behaviors. Int. J. Disaster Risk Reduct. 2024, 111, 104646. [CrossRef]
67. Yan, W.; Hu, B.; li Liu, Y.; Li, C.; Song, C. Does usage scenario matter? Investigating user perceptions, attitude and support for
policies towards ChatGPT. Inf. Process. Manag. 2024, 61, 103867. [CrossRef]
68. Farrow, E. Determining the human to AI workforce ratio–Exploring future organisational scenarios and the implications for
anticipatory workforce planning. Technol. Soc. 2022, 68, 101879. [CrossRef]
69. Smagacz-Poziemska, M.; Borowski, M.; Działek, J.; Łapniewska, Z. Post-crises (new) normality. Across social practices and
speculative fictions. Futures 2024, 155, 103292. [CrossRef]
70. OpenAI. ChatGPT-4: Optimizing Language Models for Dialogue. Available online: https://openai.com/chatgpt/ (accessed on
31 October 2024).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.

sustainability-16-11147

Uploaded by

Copyright:

Available Formats

sustainability-16-11147

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

sustainability-16-11147

Uploaded by

Copyright:

Available Formats

Article

Enhancing Organizational Resilience: Sustainable Development

Academic Editor: Sotirios Argyroudis

Received: 23 November 2024 1. Introduction

Sustainability 2025, 16, 11147. https://doi.org/10.3390/su162411147 https://www.mdpi.com/journal/sustainability

2. Background of the Research

2.1. Scenario Planning in Organizational Strategy

2.3. Challenges and Limitations in Scenario Planning

2.4. Challenges and Limitations in Scenario Planning (Old)

2.5. Innovative Approaches and the Role of Artificial Intelligence

2.6. Integrating Disaster Risk into Sustainable Development Scenarios

2.8. Challenges and Limitations of ChatGPT Use for Scenario Generation

3. The Methodology of the Research

3.1. Analyzing Dependencies: Correlation and Mutual Information Analyses

3.1.1. Correlation Analysis

3.1.2. Mutual Information with Null Model Thresholding

Null Model Approach

3.1.3. Comparison of Approaches

3.2. Factors for Evaluation

Table 1. Factors and corresponding references.

The Authors Who Also Studied

No. Parameter The Authors Who Also Studied Similar Parameters

3.3. Methodology for Scenario Generation

3.3.1. Prompt Engineering Approach

3.3.2. Approach to AI-Generated Scenarios

Micro companies (less than 10 employees)

B2.6 - Clean water and

B2.12 - Responsible B2.7 - Affordable and clean

B2.11 - Sustainable cities and B2.8 - Decent work and

In the surveyed companies, the respondents mostly experienced negative impacts of

Micro companies (less than 10 employees)

B3.1 - Risk management and

B3.6 - Ensuring sustainability in B3.4 - Attracting and mobilizing

B3.5 - Securing supply chains

Micro companies (less than 10 employees)

C1.1 - Disconnection of electricity

C1.15 - Information system

C1.14 - Deaths of employees C1.8 - A flood in the company

C1.9 - Absence of heating and

Figure 3. The frequency of negative impacts during catastrophes in companies.

4.1. Dependence Between Score Variables

C1.19 - Financial losses 1.0

4.2. Scenario Generation Using ChatGPT

4.2.1. Information Provided in Scenario Prompt Generation

concerns, including those of employees, local communities, and government agen-

4.2.2. Potential for Assessing AI-Generated Scenarios

contribute to validating the effectiveness of AI in scenario planning and identifying areas

4.2.3. Implications for Scenario Planning Using Conversational AI

4.3. ChatGPT-Generated Scenarios

4.3.1. Pessimistic Scenario (2025–2035)

Impact of Possible Disasters

Actions to Increase Organizational Resilience

• Failure to Collaborate: Does not actively seek partnerships or government assistance.

Integration of Artificial Intelligence

Addressing Sustainable Development Goals

Outcomes and Impact

4.3.2. Most Likely Scenario (2025–2035)

Impact of Possible Disasters

Actions to Increase Organizational Resilience

Integration of Artificial Intelligence