Dpia Iot

PIA, application to connected objects February 2018 edition
Contents
Foreword .............................................................................................................. 1
1 Study of the context ............................................................................................... 2
1.1 Overview of the processing ................................................................................................... 2
1.2 Data, processes and supporting assets .................................................................................. 3
2 Study of the fundamental principles ..................................................................... 6
2.1 Controls guaranteeing the proportionality and necessity of the processing ........................ 6
2.2 Controls protecting data subjects’ rights ............................................................................ 12
2.3 Assessment of compliance with the fundamental principles .............................................. 19
3 Study of data security risks ................................................................................. 20
3.1 Assessment of existing or planned controls ........................................................................20
3.2 Risk assessment: potential privacy breaches ...................................................................... 27
4 Validation of the PIA ........................................................................................... 32
4.1 Preparation of the material required for validation............................................................ 32
4.2 Formal validation of the PIA ...............................................................................................38
Appendices ................................................................................................................ 39
1. Data minimization controls..........................................................................................................39
2. Risk sources ................................................................................................................................ 40
3. Severity scale and examples of impacts ....................................................................................... 41
4. Likelihood scale and examples of threats .................................................................................... 43
5. Scales for the action plan .............................................................................................................48
6. Typology of objectives to address the risks..................................................................................48
Foreword
This document is an application of the PIA guides published by the CNIL to the specific
sector of the IoT devices (connected objects).
Performed in principle by a controller or provider, the purpose of a PIA is to build and demonstrate
the implementation of privacy protection principles so as to empower data subjects.
This is an iterative methodology, which should guarantee a reasoned, reliable use of such data during
processing.
This document is based upon the PIA method of the French Data Protection Authority (CNIL)
The methodology comprises three guides, one setting out the approach, a second containing facts that
could be used for formalizing the analysis and a third providing knowledge bases (a catalogue of
controls aimed at complying with the legal requirements and treating the risks, and examples):
These can be downloaded from the CNIL website and will be useful for completing this document:
https://www.cnil.fr/en/privacy-impact-assessments-cnil-publishes-its-pia-manual
This document is laid out like a PIA report – the deliverable of the PIA 1.
Some parts of this document [grey shaded areas] have been filled in for illustrative purposes, taking
as an example a fictional generic product comprising an interactive toy which can also be used as a
babyphone, a mobile app and an online service, for which personal data are stored by a third-party
hosting provider which calls on service providers (interactivity, analysis of uses, advertising
company).
In addition, there are notes giving advice or highlighting points to be vigilant about in the specific
context of connected objects.
Lastly, insets [beige areas] provide methodological support throughout the document to inform the
planned assessments.
1 See the WP29 guidelines on PIAs.
1
1 Study of the context
 Generally
aspects .
carried out by the project owner , with the help of a person in charge of “Data protection”
3
2
 Objective: gain a clear overview of the personal data processing operations under consideration.
1.1 Overview of the processing

 Present a brief outline of the product under consideration, its nature, scope, context,
purposes and stakes 4
 Identify the data controller and any processors.
 List the references applicable to the processing, which are necessary or must be complied
with 5, not least the approved codes of conduct (see Art. 40 of the [GDPR]) and certifications
regarding data protection (see Art. 42 of the [GDPR]) 6.
1.1.1 Product description

The template below can be used to provide a brief description of the product.
To show how to use it, it has been completed taking the example of a fictional toy, which will be used
throughout the document.
The device is a toy fitted with a microphone, a camera and basic

functional buttons (power, action, reset).
Product description
It connects via Wi-Fi and communicates with a dedicated mobile
app, hosted on a smartphone or a tablet, or with an online service.
Provide interactivity to the child, through the possibility of dialogue

with the toy (questions/answers in natural language by voice
recognition).
Processing purposes
Enable the child to communicate online (send voice messages, texts
and photos) with his/her friends and/or parents.
Feed information back to the parents (surveillance device).
Create a new category of toys for children and their parents by

Processing stakes leveraging connectivity, in keeping with the legal framework and
personal data security.
Controller The firm Fab (manufacturer)
The firms Héb (host), Int (interactive platform) and AnaPub

Processor(s)
(analysis of uses and advertising company)
2 In the business sense. This may be delegated, represented or processed by another stakeholder.
3 Such as the data protection officer for example.
4 Answer the question "What are the expected benefits (for the organization, for the data subjects, for society in
general, etc.)?"
5 Depending on the case, they will particularly be useful to demonstrate compliance with fundamental
principles, justify controls or prove that they correspond to the state of the art.
6 Other examples: security policy, sector-specific legal standards, etc.
2
1.1.2 Sector-specific references applicable to the processing7

Below you will find a table setting out the sector-specific standards applicable to your processing 8
along with the conditions for taking them into account.
Standards applicable to the processing Consideration
1.2 Data, processes and supporting assets

 Define and describe the scope in detail:
o the personal data concerned, the recipients 9 and storage durations;
o description of the processes and personal data supporting assets for the entire
personal data life cycle (from collection to erasure).
1.2.1 Data processed

Below you will find a table setting out a detailed list of the data processed and persons with access
thereto.
To show how to use it, it has been completed with the data from our example of a fictional toy.
Persons with
Personal data Categories Recipients
access thereto
Information about the user: first Authorized staff at
Common data:
name, date of birth, gender, email, The firm Héb the firms Fab and
identification data
telephone number Héb
Data entered in a third-party

Authorized staff at
app (Twitter, Facebook, etc.), Common data:
The firm Héb the firms Fab and
obtained via a link with the user identification data
Héb
account
Common data: life habits

Authorized staff
Recorded data: texts/messages, at the firms Fab
Data perceived as and Héb
sounds, images, movements, The firm Héb
sensitive: image and
temperature, humidity +
voice (enabling biometric +
processing) Authorized staff at
The firms Int
User logs on the device, mobile app and AnaPub the firms Int and
and online service AnaPub
Sensitive data (in the
meaning of the GDPR):
data relating to minors
Calculated data: answers to The firms Int Authorized staff

children's questions and and AnaPub
Common data: life habits at the firms Int
identification of interests to help
make answers more relevant + and AnaPub
7See Article 35 (8) of the [GDPR].

8For example a code of conduct, a certification, a general security policy or a PIA Framework.
9For the definition of "recipient", see Article 4(9) of the GDPR.
3
Persons with
Personal data Categories Recipients
access thereto
Sensitive data (in the The firm Héb +
meaning of the GDPR):
Analysis of uses and targeted Authorized staff at
data relating to minors
advertising the firms Fab and
Héb
1.2.2 Life cycle of data and processes

Here you need to present and describe how the product generally works, with a diagram of data flows
and a detailed description of the processes carried out.
As an example, below you will find the diagram showing how our fictional toy works.
Below you will find a table for listing in detail all the data processing operations carried out.
4
To show how to use it, it has been completed with our example of a fictional toy.
Processes Detailed description of the process
1. Open an account The user provides identification data to open his or her account
2. Capture the data Data are recorded via sensors
The data are transferred to the mobile app, directly via the device or through
3. Transfer to the mobile
the cloud servers
4. Enter the data Data are entered into the mobile app
5. Store in the mobile The data are stored in the mobile app
6. Send the data to the The data are sent to the cloud servers, via the device directly or the mobile
servers app
The interactive platform in the cloud generates the response data on the
7. Generate interactivity
basis of previous dialogues and the interests detected
The interactive data are sent back to the device, directly or through the
8. Send the data to the toy
mobile app
9. Store the data on the
The captured and calculated data are stored on the cloud servers
servers
Data analysis algorithms are run on the cloud servers to produce statistics on
10. Analyze the data
use and advertising targeting
11. Consult the cloud server Part of the captured and calculated data can be consulted via the mobile app
data or on a personal Web space
Some data can be passed on to third-party apps or posted on social media
12. Share the data
websites
1.2.3 Data supporting assets

Below you will find a table for listing in detail the data supporting assets. To show how to use it, it
has been completed with our example of a fictional toy.
IT systems 10 on which the data rely Other supporting assets 11
- Device (camera, microphone, loudspeaker, movement sensors,

temperature, humidity) - User
- User's premises
- Smartphone/tablet/computer of the user
- Mobile app/browser - Premises of Fab and Héb
- Staff at Fab and Héb
- Wi-Fi network
- Internet - Premises of Int and AnaPub
- Staff at Int and AnaPub
- Cloud servers of Héb, Int and AnaPub
 NB: the whole of Part 1 "Context" must be read through by the DPO to make sure that it is exhaustive
and properly reflects the way things really are.
This is all the more necessary given that this part describes the key aspects and notions on which the
following chapters are based.
10 Can be broken down into hardware (and electronic data media), software and computer channels.
11 Can be broken down into people, paper documents and paper transmission channels.
5
2 Study of the fundamental principles
 Generally performed by the project owner and then assessed by a person in charge of "Data
Protection".
 Objective: build the system that ensures compliance with privacy principles.
The fundamental principles bearing on privacy protection which must be taken on

board are as follows: specified and explicit data collection purposes, lawfulness of the processing,
data minimization, data quality, limited storage durations, information for the data subjects,
obtaining their consent, possibility of accessing their data directly, portability of their data, possibility
of rectifying and erasing their data at their request, possibility of objecting to or restricting the
processing, regulation of processors and transfer of data outside the European Union.
 Explain and justify the choices made and describe the controls selected (existing or
planned) to comply with these legal requirements (it is necessary to explain how it is
intended to implement them).
 Check that improving the way in which each point is planned, clarified and justified, pursuant
to the [GDPR], is either not necessary or not possible.
 Where applicable, review their description or propose additional controls.
 Note: In Para. 2.3 below, you will find a table for summarizing the justifications for all these points
and for recording their assessment and any corrective controls.
2.1 Controls guaranteeing the proportionality and necessity of the

processing
2.1.1 Purposes: specified, explicit and legitimate12

Below you will find a table for setting out in detail the data processing purposes and for justifying their
legitimacy 13.
Purposes Legitimacy
 Note: remember to explain the purposes of sharing with third parties, in particular for advertising
and "partner offers", as well as the data processing purposes for improving the service.
 Note: remember to explain the specific conditions under which the processing will take place,
particularly by clarifying data matching where applicable.
12
See Article 5.1 (b) of the [GDPR].
13On the legitimacy of the purpose, see opinion WP 203 of the Article 29 Data Protection Working Party -
http://ec.europa.eu/justice/data-protection/article-29/documentation/opinion-
recommendation/files/2013/wp203_en.pdf.
6
 NB 14 : on account of a child's general vulnerability and the fact that personal data must be processed
fairly and lawfully, the controllers of a processing operation targeting children must comply even
more strictly with the principles of purpose limitation.
More particularly, the controllers must not use the child's data for profiling purposes (e.g. for
targeted advertising), whether directly or indirectly, insofar as it is not possible for a child to
understand the implications of this, and it therefore goes beyond what can be considered fair
processing.
2.1.2 Basis: lawfulness of processing, prohibition of misuse15

Below you will find the list of lawfulness criteria. Processing shall be lawful only if and to the extent
that at least one of the following applies:
Lawfulness criteria Applicable Justification
The data subject has given consent 16 to the processing of his or

her personal data for one or more specific purposes
Processing is necessary for the performance of a contract to

which the data subject is party or in order to take steps at the
request of the data subject prior to entering into a contract
Processing is necessary for compliance with a legal obligation

to which the controller is subject
Processing is necessary in order to protect the vital interests of

the data subject or of another natural person
Processing is necessary for the performance of a task carried

out in the public interest or in the exercise of official authority
vested in the controller
Processing is necessary for the purposes of the legitimate

interests pursued by the controller or by a third party, except
where such interests are overridden by the interests or
fundamental rights and freedoms of the data subject which
require protection of personal data, in particular where the
data subject is a child 17
 Note: where processing is carried out in accordance with a legal obligation or where processing is
necessary for the performance of a task carried out in the public interest or in the exercise of official
authority, clarify in the justification the legal basis for the processing in Union law or the law of the
Member State to which the controller is subject.
 Note: there can be several types of basis for a processing operation: for example, a contract associated
with the purchase of a product for using it for its primary purpose and consent for its secondary
purposes (improving the service, marketing, etc.) which will be obtained when the product is
activated.
 NB: Where the processing for a purpose other than that for which the personal data have been
collected is not based on the data subject's consent or on a Union or Member State law, the controller
14 See the WP29 opinion 02/2013 on apps on smart devices.

15
See article 6 of the [GDPR].
16 With regard to obtaining the data subject's consent and informing the latter, see Chapter 2.2.
17 This point shall not apply to processing carried out by public authorities in the performance of their tasks
7
shall, in order to ascertain whether processing for another purpose is compatible with the purpose
for which the personal data are initially collected, take into account, inter alia:
 any link between the purposes for which the personal data have been collected and the
purposes of the intended further processing;
 the context in which the personal data have been collected, in particular regarding the
relationship between data subjects and the controller;
 the nature of the personal data, in particular whether special categories of personal data are
processed, or whether personal data related to criminal convictions and offences are
processed 18;
 the possible consequences of the intended further processing for data subjects;
 the existence of appropriate safeguards, which may include encryption or pseudonymization.
2.1.3 Data minimization: adequate, relevant and limited 19

It is important to reduce the severity of the risks by minimizing the number of personal data that will
be processed, by limiting such data to what is strictly necessary for the purposes for which they are
processed (otherwise they should not be collected). Then, it also becomes possible to minimize the
data themselves, via controls aimed at reducing their sensitivity (see Appendix 1 - List of data
minimization controls).
Below you will find a table for listing the data processed, reduced to what is strictly necessary,
alongside the justification of the need and any additional minimization controls.
Justification of the
Details about the data Minimization
Data types Data categories need and relevance of
processed controls
the data
First name, date of Details necessary for No surname

birth, email, telephone creating a profile for
number, link with a communicating
Replacing the date
social media account
of birth with the
Civil status, identity,
age or age group
identification data
Separate storage of
identifying data in
an encrypted base
Texts/messages, sounds, Aspects that are part
Common images, movements, of the
data Personal life (living temperature, humidity communication
habits, marital Answers to children's features
status, excluding questions and
sensitive or identification of
dangerous data, etc.) interests to help make
answers more relevant,
targeted advertising
Professional life Not collected
(résumé, education
and professional
training, awards,
etc.)
18 See Articles 9 and 10 of the [GDPR].

19
See Article 5.1 (c) of the [GDPR].
8
Justification of the
Details about the data Minimization
Data types Data categories need and relevance of
processed controls
the data
Economic and Not collected

financial
information
(income, financial
situation, tax
situation, etc.)
Application traces Required for security Pseudonymization
Connection data (IP
Technical logs reasons and to check for statistical use
addresses, events
compliance with the
logs, etc.)
ST&Cs
Smartphone location Not necessary Removal of
Location data
integrated in the photos location
(travels, GPS data,
(if option is activated) information before
GSM data, etc.)
photos are sent
Social security Not collected
number
Data Raw data: voice and Aspects that are part
perceived photographs of the
as Biometric data
communication
sensitive features
Bank data Not collected
Opinions bearing on Not collected but can Aspects that are part
philosophy, politics, appear directly or of the
religion, trade union indirectly in the text, communication
involvement, audio and video data features
Sensitive sexuality, health
data in the data, racial or ethnic
meaning origin, data
of the concerning health or
[DP-Act] 20 sexuality
Offences, Not collected

convictions, security
measures
 Notes: remember to clearly justify the collection of certain data (location, date of birth, age, weight,
etc.) and clearly distinguish between anonymous and pseudonymous data.
 Tip: avoid free text input fields (like "comments" fields), because of the risk of users noting down
there information that does not comply with the minimization principles. Preference should
therefore be given to scroll-down list type fields. If free-form text fields cannot be avoided, users'
awareness must be raised in how to use such fields, with regard to the standard terms & conditions
for service and the law (no offensive words, no undeclared sensitive data, etc.).
 NB: for processing of minors' details, the data are considered overall to be sensitive pursuant to the
[GDPR].
20 Also see Articles 9 & 10 of the [GDPR]. Restrictions of use and special formalities are to be taken into account.
9
 NB 21 : on account of a child's general vulnerability and the fact that personal data must be processed
fairly and lawfully, the controllers of a processing operation targeting children must comply even
more strictly with the principles of data minimization and purpose limitation.
The data controllers should also specifically refrain from any collection of data relating to the parents
or family members of the child user, such as financial information or information about special
categories of information, such as medical data.
2.1.4 Quality of data: accurate and kept up-to-date22

Below you will find a table for setting out in detail the data quality compliance controls, carried out
on the device, the mobile app and the personal account, as well as a justification on the arrangements
for or impossibility of implementing them.
Personal
Data quality controls Device Mobile app Justification
account
Regular checks of the accuracy of the user's

personal data
Invitation for the user to check and, where

necessary, update his or her data
Traceability of data amendments
2.1.5 Storage durations: limited23

A storage duration must be defined for each type of data and justified by the legal requirements and/or
processing needs. Thus a distinction is made between common data and archived data, to which
access will be limited to only the stakeholders concerned.
An erasure mechanism must be implemented to archive common data or purge archived data at the
end of their storage duration. Functional traces will also have to be purged, as will technical logs which
may not be stored indefinitely.
Notes: By reducing the amount of available and processed data, archiving and purging help to limit
 the impacts in the event of theft or accidental dissemination of the database.
In order to ensure that these storage durations are effective, it is recommended to set up an automatic
mechanism based on the date on which the data are created or last used.
 NB: For sensitive data and high-risk data, use should be made of secure erasure tools that make the
data irretrievable.
21 See WP29 opinion 02/2013 on apps on smart devices.

22
See Article 5.1 (d) of the [GDPR]). The quality requirement also concerns the link between the data identifying
the people and the data concerning them.
23
See Article 5.1 (e) of the [GDPR], unless there is another legal obligation calling for longer storage periods).
10
The storage durations, their justification and purge mechanisms can be presented in the table below.
Justification of the Erasure mechanism at the end
Data types Storage duration
storage duration of the storage duration
Common data
Archived data
Functional traces
Technical logs
11
2.2 Controls protecting data subjects’ rights
2.2.1 Information for the data subjects (fair and transparent processing)24
If the processing benefits from an exemption from the right to information, as provided for in Articles
12, 13 & 14 of the [GDPR], you will need to justify this below.
Exemption from having to inform the data subjects Justification
Otherwise, below you will find a list of controls intended to provide information to users (or their
parents) 25.
You need to describe how they are implemented (preferably by attaching screenshots and document
extracts) on the device, mobile app and personal account, and justify the arrangements for or
impossibility of implementing them.
Personal
Controls for the right to information Device Mobile app Justification
account
Presentation, when the device is activated, of the

terms & conditions for use/confidentiality
Possibility of accessing the terms & conditions
for use/confidentiality after activation
Legible and easy-to-understand terms
Existence of clauses specific to the device
Detailed presentation of the data processing

purposes (specified objectives, data matching
where applicable, etc.)
Detailed presentation of the personal data
collected
Presentation of any access to the identifiers of
the device, the smartphone/tablet or computer,
specifying whether these identifiers are
communicated to third parties
Presentation of the user's rights (consent
withdrawal, data erasure, etc.)
Information for the user if the app is likely to run
in the background
Information on the secure data storage method,
particularly in the event of sourcing
Information on protection of access to the device
Arrangements for contacting the company

(identity and contact details) about
confidentiality issues
24
See Articles 12, 13 & 14 of the [GDPR].
25See the CNIL's website: "Editeurs de sites pour enfants : n’oubliez pas vos obligations !" (Publishers of
websites for children: remember your obligations!).
12
Personal
Controls for the right to information Device Mobile app Justification
account
Information on the possibility of defining

directives concerning the post-mortem fate of
data
Where applicable, information for the user on
any change concerning the data collected, the
purposes and confidentiality clauses
Regarding transmission of data to third parties:
- detailed presentation of the purposes of

transmission to third parties
- detailed presentation of the personal data

transmitted
- indication of the identity of third-party bodies
 NB: in the event that data are transmitted to third-party bodies in relation to the data controller
(subsidiaries, members, intra-group, partners, etc.), it is necessary to supply the list of recipients (in
a dedicated information section), clarifying the data categories transmitted and the transfer purpose,
and providing a hyperlink to the data protection policy of the respective recipients. An internal
process must also be planned so as to be able to update this list in the event of changes.
 NB 26: App developers, in collaboration with app stores and operating system and device
manufacturers, should present the relevant information in a simple manner, in age specific language
adapted to young children – possibly via a sound message.
 Recommendation: place an information “QR Code" on the device and make the users (or their
parents) aware of their responsibility to inform third parties that their data are likely to be collected
(e.g. the other children conversing with the device or featured on the shared photos).
2.2.2 Obtaining consent, where applicable: express

Where processing is based on consent, the controller shall be able to demonstrate that the data subject
has consented. The data subject must be able to withdraw his/her consent easily at any time 27.
Where the lawfulness of the processing 28 is based on consent, below you will find a list of controls
intended to ensure that users' (or their parents') consent has been obtained 29, that there has been a
reminder and confirmation of their consent, and the settings associated with the latter have been
maintained.
You will have to describe their implementation on the device, the mobile app and the personal
account, as well as a justification on the arrangements for or impossibility of implementing them.

27
See Articles 7 & 8 of the [GDPR].
28 Concerning the lawfulness of processing, see chapter 2.1.
29 See on CNIL website: "Editeurs de sites pour enfants : n’oubliez pas vos obligations !" (Publishers of websites
for children: remember your obligations!).
13
Personal
Controls for obtaining consent Device Mobile app Justification
account
Express consent during activation
Consent segmented per data category or

processing type
Express consent prior to sharing data with other

users
Consent presented in an intelligible and easily

accessible form, using clear and plain language
adapted to the target user (particularly for
children)
Obtaining parents' consent for minors under 13

years of age
For each new user, consent must once again be

obtained
After a long period without use, the user must be

asked to confirm his/her consent
Where the user has consented to the processing

of special data (e.g. his/her location), the
interface clearly indicates that said processing
takes place (icon, light)
Where the user changes device, smartphone or

computer, reinstalls the mobile app or deletes
his/her cookies, the settings associated with
his/her consent are maintained
 NB 30: the GDPR has strengthened the legal basis regarding consent for any direct provision of
information society services to minors, and the burden of proof (unambiguous) lies with the data
controller or processor.
In practice, the consent of the holder of parental responsibility is required for children under 16 years
of age, with Member States able to determine a lower age, as long as this is not below 13 years of age.
The data controller must make reasonable efforts to check that the holder of parental responsibility
has indeed consented, in view of the technological means available.
 NB 31: When consent can legally be obtained from a minor, and the app is intended to be used by a
child or a minor, the data controller should pay attention to the minor’s potentially limited
understanding of and attention for information about data processing.
App developers, in collaboration with app stores and operating system and device manufacturers,
should present the relevant information in a simple manner, in age specific language.
30 See article 8 of [GDPR].

14
2.2.3 Exercising rights of access32 and to data portability33

Where the processing benefits from an exemption from the right of access, as provided for in Articles
15 of the [GDPR], you will need to justify this below.
Exemption from the right of Arrangements for responding to
Justification
access the data subjects
Otherwise, below you will find a list of the controls intended to ensure users' (or their parents') right
of access to all personal data concerning them.
Personal
Controls for the right of access Device Mobile app Justification
account
Possibility of accessing all of the user's personal

data, via the common interfaces
Possibility of securely consulting the traces of use

associated with the user
Possibility of downloading an archive of all the

personal data associated with the user
Lastly, where the right to data portability applies to processing pursuant to Article 20 of the [GDPR],
you will have to describe its implementation below.
Personal
Controls for the right to data portability Device Mobile app Justification
account
Possibility of retrieving, in an easily reusable

format, personal data provided by the user, so as
to transfer them to another service
32
See Article 39 of the [DP-Act] and Article 15 of the [GDPR].
33 See Article 48 of Act 2016-1321 of 7 October 2016 for a digital Republic and Article 20 of the [GDPR].
15
2.2.4 Exercising the rights to rectification and erasure34

Where the processing benefits from an exemption from the right to rectification and erasure, as
provided for by Article 17 of the [GDPR] you will need to justify this below.
Exemption from the rights to Justification Arrangements for responding to
rectification and erasure the data subjects
Otherwise, below you will find a list of controls intended to ensure the right to rectification or erasure
of data of users (or their parents 35) who request this.
Controls for the rights to rectification and Personal
Device Mobile app Justification
erasure account
Possibility of rectifying personal data
Possibility of erasing personal data
Indication of the personal data that will

nevertheless be stored (technical requirements,
legal obligations, etc.)
Implementing the right to be forgotten for

minors
Clear indications and simple steps for erasing

data before scrapping the device
Advice given about resetting the device before

selling it
Possibility of erasing the data in the event the

device is stolen
 NB 36: The data controller has one month in which to erase the data or respond to the data subject;
beyond this time limit, the data subject can refer the case to its Data Protection Authority (the CNIL
in France). There are exceptions, particularly in the event the information published is necessary for
the freedom of information, on the grounds of public interest or to comply with a legal obligation.
An Internet user under 18 years of age at the time of publication or creation of an online account can
directly and without the need for an explanation, ask the website to erase data concerning him/her
at the earliest possible opportunity.
34
See Articles 40 & 41 of the [DP-Act] and Articles 16, 17 & 19 of the [GDPR].
35 See on CNIL website: "Editeurs de sites pour enfants : n’oubliez pas vos obligations !" (Publishers of websites
for children: remember your obligations!).
36 See Act 2016-1321 of 7 October 2016 for a digital Republic amending Article 40 of the [DP-Act], which rounds
off the "right to be forgotten" as provided for by Article 17 of the [GDPR].
16
2.2.5 Exercising the rights to restriction of processing and to object37

Where the processing benefits from an exemption from the right to restriction and to object, as
provided for by Article 21 of the [GDPR], you will need to justify this below.
Exemption from the rights to Arrangements for responding to
Justification
restriction and to object the data subjects
Otherwise, below you will find a list of controls intended to ensure the right to object and to restriction
either concerning the different purposes or the whole of a processing operation.
Controls for the rights to restriction and to Personal
Device Mobile app Justification
object account
Existence of "Privacy" settings
Invitation to change the default settings
"Privacy" settings accessible when activating

the device
"Privacy" settings accessible after activating

the device
Existence of a parental control system for

children under 13 years of age
Existence of a system allowing the user to ask

for the processing to be restricted
Existence of technical means for the data

controller to lock access to and use of the data
subject to the restriction
Possibility of deactivating some of the device's

features (microphone, Web browser, etc.)
Existence of alternative apps for accessing the

device
Possibility of objecting to the mobile app

running in the background
Compliance in terms of tracking (cookies,

advertising, etc.)
Exclusion of children under 13 years of age

from automated profiling
Effective exclusion of processing the user's

data in the event consent is withdrawn
37
See Articles 18 & 21 of the [GDPR].
17
 Note: the right to restriction allows the data subject to call for processing of his/her data to be
“frozen”, as a protective control while its legitimacy is being checked, for example.
2.2.6 Processors: identified and governed by a contract38

A processing contract must be signed with each processor, setting out all of the aspects stipulated in
Art. 28 of the [GDPR]: duration, scope, purpose, documented processing instructions, prior
authorization where a processor is engaged, provision of any documentation providing evidence of
compliance with the [GDPR], prompt notification of any data breach, etc.
Below you will find a table for setting out in detail the contracts for each of the processors.
Contract Compliance with
Processor's name Purpose Scope
reference Art.28
2.2.7 Transfers: compliance with the obligations bearing on transfer of data outside the
European Union39
Below you will find a table for setting out in detail the geographic storage location of the device, mobile
app and personal account data in the cloud.
Depending on the country in question, you will have to justify the choice of remote hosting and
indicate the legal supervision arrangements implemented in order to ensure adequate protection of
the data subject to a cross-border transfer.
Country Justification and supervision
recognized as Other (standard contractual
European providing country clauses, internal corporate
Data storage location France
Union adequate regulations)
protection by
the EU
Device data
Mobile app data
Personal account data

39
See Articles 44 to 50 of the [GDPR].
18
2.3 Assessment of compliance with the fundamental principles

Below you will find a table for summarizing, for each point concerning compliance with the legal
requirements, the way in which it is applied in the processing.
The last two columns are for use by the assessor:
 Acceptable/can be improved on?
The assessor must estimate whether the controls allow compliance with the fundamental principles.
 Corrective controls:
Where applicable, he shall indicate any additional controls that would prove necessary.
Controls guaranteeing the Acceptable/ca

proportionality and necessity of the Justification n be improved Corrective controls
processing on?
Purposes: specified, explicit and

legitimate
Basis: lawfulness of processing,

prohibition of misuse
Data minimization: adequate,

relevant and limited
Quality of data: accurate and kept

up-to-date
Storage durations: limited

Acceptable/ca
Controls to protect the rights of data
Justification n be improved Corrective controls
subjects
on?
Information for the data subjects

(fair and transparent processing)
Obtaining consent
Exercising the rights of access and to

data portability
Exercising the rights to rectification

and erasure
Exercising the rights to restriction of

processing and to object
Processors: identified and governed

by a contract
Transfers: compliance with the

obligations bearing on transfer of
data outside the European Union
19
3 Study of data security risks 40

A risk is a hypothetical scenario that describes a feared event and all the threats that allow this to
occur. More specifically, it describes:
 how risk sources (e.g.: an employee bribed by a competitor)
 could exploit the vulnerabilities of supporting assets (e.g.: the file management system that
allows the manipulation of data)
 in a context of threats (e.g.: misuse by sending emails)
 and allow feared events to occur (e.g.: illegitimate access to personal data)
 on personal data (e.g.: customer file)
 thus generating impacts on the privacy of data subjects (e.g.: unwanted solicitations, feelings
of invasion of privacy, personal or professional problems).
3.1 Assessment of existing or planned controls
 Generally
aspects .
performed by the prime contractor
42
41, then assessed by a person in charge of “Data security”
 Objective: gain a good understanding of the controls that contribute to security.
 Identify or determine the existing or planned controls (already undertaken), which can
take three different forms:
1. controls bearing specifically on the data being processed: encryption,
anonymization, partitioning, access control, traceability, etc.;
2. general security controls regarding the system in which the processing is
carried out: operating security, backups, hardware security, etc.;
3. organizational controls (governance): policy, project management, personnel
management, management of incidents and breaches, relations with third parties, etc.
 Check that improving each control and its description, pursuant to best security practice, is
either not necessary or not possible.
 Where applicable, review their description or propose additional controls.
 Notes: The security control categories below correspond to the CNIL's recommended good
practices 43.
You will also need to take account of the sector-specific standards applicable to your processing 44
(general security policy, PIA Framework, code of conduct, etc.).
 Note: In Para. 3.1.4 below, you will find a table for summarizing the implementation of all these
controls and for recording their assessment and any corrective controls.

41 This may be a delegate, representative or processor.
42 Chief information security officer or other.
43 See the "Guide on security of personal data" by the CNIL.
44 See Article 35 (8) of the [GDPR].
20
3.1.1 Controls bearing specifically on the data being processed
Encryption
Describe here the means implemented for ensuring the confidentiality of data stored (in
the database, in flat files, backups, etc.), as well as the procedure for managing encryption keys
(creation, storage, change in the event of suspected cases of data compromise, etc.).
Describe the encryption means employed for data flows (VPN, TLS, etc.) implemented in the
processing.
 Notes: think about Wi-Fi security (encryption, Wi-Fi password storage).

Think about securing the certificates stored on the device or smartphone, used for authenticating and
encrypting connections.
Anonymization
Indicate here whether anonymization mechanisms are implemented, which ones and for what
purpose.
 Note: remember to clearly distinguish between anonymous and pseudonymous data.
Data partitioning (in relation to the rest of the information system)

Indicate here if processing partitioning is planned, and how this is carried out.
Logical access control

Indicate here how users' profiles are defined and attributed.
Specify the authentication means implemented 45.
Where applicable, specify the rules applicable to passwords (minimum length, required
characters, validity duration, number of failed attempts before access to account is locked, etc.).
 Notes: think about the security of the user's password, whether on the device, smartphone or in the
cloud. Passwords must be stored properly hashed by a robust algorithm with a salt applied
beforehand.
Think about protecting access to the app on the smartphone with a specific password.
Think about securing peering between the device, mobile app and personal account.
Think about protecting the data, including metadata (including Exif) and technical traces in the event
of direct access via physical connection to the device or smartphone.
45See the CNIL deliberation no. 017-012 of 19 January 2017 on the adoption of a recommendation relating to
passwords.
21
Traceability (logging)
Indicate here whether events are logged and how long these traces are stored for.
Integrity monitoring
Where applicable, indicate here whether mechanisms are implemented for integrity monitoring of
stored data, which ones and for what purpose.
Specify which integrity control mechanisms are implemented on data flows.
Archiving
Where applicable, describe here the processes of archive management (delivery, storage,
consultation, etc.) under your responsibility. Specify the archiving roles (offices of origin,
transferring agencies, etc.) and the archiving policy.
State if data may fall within the scope of public archives.
Paper document security

Where paper documents containing data are used during the processing, indicate here how they are
printed, stored, destroyed and exchanged.
3.1.2 General security controls regarding the system in which the processing is carried
out
The following controls generally concern the security of the whole body. They can particularly be
formally documented in a cybersecurity policy (PSSI) or equivalent.
Operating security
Describe here how the software updates (operating systems, applications, etc.) and application
of security patches are carried out.
 Note: think about the possibilities of updating the device.
Managing workstations and clamping down on malicious software

Describe here the controls implemented on workstations (automatic locking, firewall, etc.) and state
whether an antivirus software is installed and updated at regular intervals on the workstations.
Website security
Indicate here whether ANSSI's "Recommendations for securing websites" have been implemented.
22
Backups
Indicate here how backups are managed. Clarify whether they are stored in a safe place.
Maintenance
Describe here how physical maintenance of hardware is managed, and state whether this is
contracted out.
Indicate whether the remote maintenance of apps is authorized, and according to what
arrangements.
Specify whether defective equipment is managed in a specific manner.
Security of computer channels (networks)

Indicate here the type of network on which the processing is carried out (isolated, private or
Internet).
Specify which firewall system, intrusion detection systems or other active or passive devices are in
charge of ensuring network security.
Monitoring
Indicate here whether real-time monitoring of local network is implemented and with what means.
Indicate whether monitoring of hardware and software configurations is carried out and by what
means.
Physical access control

Indicate here how physical access control is carried out regarding the premises accommodating the
processing (zoning, escorting of visitors, wearing of passes, locked doors and so on).
Indicate whether there are warning procedures in place in the event of a break-in.
Hardware security
Indicate here the controls bearing on the physical security of servers and workstations
(secure storage, security cables, confidentiality filters, secure erasure prior to scrapping, etc.).
Avoiding sources of risk

Indicate here whether the implantation area is subject to environmental disasters (flood zone,
proximity to chemical industries, earthquake or volcanic zone, etc.).
Specify if dangerous products are stored in the same area.
Protecting against non-human sources of risks

Describe here the means of fire prevention, detection and fighting.
Where applicable, indicate the means of preventing water damage.
Also specify the means of power supply monitoring and relief.
3.1.3 Organizational controls (governance)

23
Organization
Indicate if the roles and responsibilities for data protection are defined.
Specify whether a person is responsible for the enforcement of privacy laws and regulations.
Specify whether there is a monitoring committee (or equivalent) responsible for the guidance
and follow-up of actions concerning the protection of privacy.
Policy (management of rules)

Indicate whether there is an IT charter (or equivalent) on data protection and the correct use of IT
resources.
Risk management
Indicate here whether the privacy risks posed by new treatments on data subjects are assessed,
whether or not it is systematic and, if applicable, according to which method.
Specify whether an organization-level mapping of privacy risks is established.
Project management
Indicate here whether device tests are performed on non-real/anonymous data.
Management of incidents and data breaches

Indicate here whether IT incidents are subject to a documented and tested management procedure.
Personnel management
Indicate here what awareness-raising controls are carried out with regard to a new recruit.
Indicate what controls are carried out when persons who have been accessing data leave their job.
Relations with third parties

Indicate here, for processors requiring access to data, the security controls and arrangements
carried out as regards such access.
Supervision
Indicate here whether the effectiveness and adequacy of privacy controls are monitored.
24
3.1.4 Assessment of security controls

Below you will find a table for summarizing how each of the security controls recommended by the
CNIL is implemented or for justifying why it is not.
The last two columns are for use by the assessor:
The assessor must determine whether the controls comply with the CNIL's recommended good practices.
Where applicable, he shall indicate any additional controls that would prove necessary.
Acceptable/c
Controls bearing specifically on the data Implementation or an be
Corrective controls
being processed justification why not improved
on?
Encryption
Anonymization
Data partitioning (in relation to the rest of
the information system)
Logical access control
Traceability (logging)
Integrity monitoring
Archiving
Paper document security
Acceptable/c
General security controls regarding the
Implementation or an be
system in which the processing is carried Corrective controls
justification why not improved
out
on?
Operating security
Managing workstations and clamping
down on malicious software
Website security
Backups
Maintenance
Security of computer channels (networks)
Monitoring
Physical access control
Hardware security
Avoiding sources of risk
Protecting against non-human sources of
risks
25
Acceptable/c
Implementation or an be
Organizational controls (governance) Corrective controls
justification why not improved
on?
Organization
Policy (management of rules)
Risk management
Project management
Management of incidents and data
breaches
Personnel management
Relations with third parties
Supervision
26
3.2 Risk assessment: potential privacy breaches
 Generally
aspects.
performed by the project owner, then assessed by a person in charge of “Data protection”
 Objective: gain a good understanding of the causes and consequences of risks.
 For each feared event (illegitimate access to personal data 46, unwanted change of personal
data 47, and disappearance of personal data48) :
o determine the potential impacts 49 on data subjects' privacy if it occurred 50 ;
o estimate its severity, particularly depending on the prejudicial nature of the potential
impacts and, where applicable, controls likely to modify them;
o Identify the threats51 to personal data supporting assets that could lead to this feared
event 52 and the risk sources 53 that could cause it;
o estimate its likelihood, particularly depending on the level of vulnerabilities of
personal data supporting assets, the level of capabilities of the risk sources to exploit
them and the controls likely to modify them;
 Determine whether the risks identified in this way 54 can be considered acceptable in view of
the existing or planned controls (already undertaken).
 If not, propose additional controls and re-assess the level of each of the risks in view of the
latter, so as to determine the residual risks 55.
 NB: since the existing or planned controls (already undertaken) are taken on board in the risk
assessment, before moving on to Part 3.2, the controls identified in Para. 2 (legal) and Para. 3.1
(security) must first have been assessed to ensure that their list is exhaustive and properly reflects the
way things really are.
 NB: any corrective controls suggested by the assessor in Paras 2.3 and 3.1.4 must be taken into account
during the residual risk calculation in Paras 3.2.1, 3.2.2 and 3.2.3, at the same time as the corrective
controls specific to each of the risks.
All of the corrective controls will be set out in the action plan in Para. 4.1.
46 They are known to unauthorised persons (data confidentiality breach).

47They are altered or changed (breach of personal data integrity).
48They are not or no longer available (breach of personal data availability).
49 See Appendix 3 – Severity scale and examples of impacts.
50 Answer the question "What do we fear might happen to data subjects?"
51 See Appendix 4 – Likelihood scale and examples of threats.
52 Answer the question "How might this happen?"
53 See Appendix 2 – Risk sources.
54 A risk is based upon a feared event and all threats that would make it possible.
55 Risks that remain after the controls have been implemented.
27
3.2.1 Illegitimate access to personal data
Assessment of the risk

Below you will find a table for noting down the result of the analysis of this risk.
Main controls
Main risk Main potential
Risk Main threats 57 reducing the severity Severity 60 Likelihood 61
sources 56 impacts 58
and likelihood 59
Rogue Minimization
acquaintances Storage durations
Consequences
Rogue of the Logical access control
neighbor disclosure of Stream encryption
Data potentially
theft/consultation (SSL)
sensitive
Rogue on the server information Hardware
Illegitimate employee (discrimination, authentication
access to threats, attacks, Private cloud
Account theft (via Significant Maximum
personal loss of Logical access control
Authorized a smartphone)
data employment,
third-party Employee clearance
company loss of access to
Recovery of a services, etc.) Access logging
scrapped device Log audits
Phishing
Hacker Notification of data
targeting a Targeted
advertising subject violations and
user or one of recommendation of
the suitable preventive
companies controls
Describe here a few representative scenarios of the risk of illegitimate data access, by spelling out
the sources, threats and impacts.
Below you will find an illustration based on our example of a fictional toy:
Data could be stolen by an employee with a profit motive or malicious intent, consulted by family
or friends taking over the account via the smartphone, or retrieved on a scrapped device by
neighbors or a hacker with a view to characterizing a situation bearing on the data subjects' private
life.
56 Relevant sources for this risk, among those identified in the context of the processing (see Appendix 2 – Risk
sources).
59 Controls among those identified in Para. 2 (legal) and Para. 3.1 (security).
28
Assessment of residual risks

The assessor must determine whether the existing or planned controls (already undertaken)
sufficiently reduce this risk for it to be deemed acceptable.
Where applicable, he shall indicate here any additional controls that would prove necessary.
 Residual risks:
The assessor will indicate here the residual risk for the processing once the aforementioned
additional controls have been implemented, by determining the severity and likelihood in view of
these controls.
Severity: Likelihood:
 NB: an additional control carried out to deal with one of the risks can also have a positive or negative
effect on the other risks.
Below you will find an illustration based on our example of a fictional toy:
 Can be improved on:
The planned controls do not sufficiently reduce this risk for it to be deemed acceptable.
- carry out encryption of the data stored in the base;
- inform the user of the good practices to follow when scrapping the device;
- supply a charter on using IT equipment and a confidentiality undertaking for employees.
 Residual risks:
Data could be consulted by family or friends, taking over the account via the smartphone.
Severity: Significant Likelihood: Negligible
29
3.2.2 Unwanted change of data

Below you will find a table for noting down the result of the analysis of this risk. To show how to use
it, it has been completed with the data from our example of a fictional toy.
Main risk Main potential Main controls reducing the
Risks Main threats Severity Likelihood
sources impacts severity and likelihood
Negligent
or rogue
user Backup of the cloud server
/family
member Stream encryption (SSL)
/friend Hardware authentication
Identity theft Private cloud
Rogue Logical access control
Unwanted neighbor Alteration of
change of data on the Deterioration in Employee clearance Limited Limited
data server the service Access logging
Negligent
quality Log audits
or rogue
employee Notification of data subject
violations and
recommendation of
Hacker suitable preventive
targeting controls
one of the
companies
Describe here a few representative scenarios of the risk of an unwanted change of data by spelling
out the sources, threats and impacts.

 Residual risks:
these controls.
30
3.2.3 Disappearance of data

Below you will find a table for noting down the result of the analysis of this risk. To show how to use
it, it has been completed with the data from our example of a fictional toy.
Main risk Main potential Main controls reducing
Risks Main threats Severity Likelihood
sources impacts the severity and likelihood
Negligent
or rogue
user
Backup of the cloud server
/family
member
/friend Need to recreate Private cloud
Erasure of data a user account Physical protection of the
Negligent (via the app or cloud servers
or rogue server) Loss of history Maintenance
Disappearance employee Deterioration of and personal Temporary on-premises
Limited Limited
of data servers service settings data retention
Hacker Physical Logical access control
targeting a damage to the Deterioration in Employee clearance
user or device the service Strong authentication of
one of the quality employees
companies
Access logging
Warranty for the device
Damage at
one of the
companies
Describe here a few representative scenarios of the risk of data disappearing, by spelling out the
sources, threats and impacts.

 Residual risks:
these controls.
31
4 Validation of the PIA
 Generally
aspects.
performed by the controller, with the help of a person in charge of "Data Protection"
 Objective: decide whether or not to accept the PIA in light of the study's findings.
4.1 Preparation of the material required for validation

 Consolidate and present the study's findings:
1. prepare a visual presentation of the controls selected to ensure compliance with
the fundamental principles, depending on their compliance with the [GDPR] (e.g.:
conditional on improvement or considered compliant);
2. prepare a visual presentation of the controls selected to contribute to data
security, depending on their compliance with best security practice (e.g.: conditional
on improvement or considered compliant);
3. visually map the risks (initial and residual where applicable 62) depending on their
severity and likelihood;
4. draw up an action plan based on the additional controls identified during the
previous steps: for each control, determine at least the person responsible for its
implementation, its cost (financial or in terms of workload) and estimated timeframe.
 Formally document the consideration of stakeholders:

1. the advice of the person in charge of "Data Protection" aspects 63;
2. the view of data subjects or their representatives64.
 Note: The spaces for noting down the assessment of controls and risks are inserted directly into the
previous parts, as near to the aspects to be assessed as possible.
All of the parts must be assessed before deciding on whether the PIA can be validated or not.
62 Risks that remain after the controls have been implemented.

63 See Article 35 (2) of the [GDPR]
64 See Article 35 (9) of the [GDPR]
32
4.1.1 Mapping of compliance with the fundamental principles

Below you will find a graph illustrating the controls bearing on compliance with the fundamental
principles, with a conformity value attributed to each on the basis of its assessment in Para. 2.3.
If the additional controls are implemented correctly, compliance with the fundamental principles
could be illustrated as follows:
Compliance with the fundamental principles

Existing or planned With corrective controls
Purposes: specified, explicit

and legitimate
Transfers: compliance with Basis: lawfulness of
the obligations bearing on processing, prohibition of
transfer of data outside… misuse
Data minimization:
Processors: identified and
adequate, relevant and
governed by a contract
limited
Exercising the rights to

Quality of data: accurate
restriction of processing and
and kept up-to-date
to object
Exercising the rights to

Storage durations: limited
rectification and erasure
Exercising the rights of Information for the data

access and to data subjects (fair and
portability transparent processing)
Obtaining consent
Graph scale:
0. Non applicable
1. Can be improved on
2. Acceptable
3. Good practices
33
4.1.2 Mapping of compliance with good security practices

Below you will find a graph illustrating the good security practices, with a conformity value attributed
to each on the basis of its assessment in Para. 3.1.4.
If the additional controls are implemented correctly, compliance with the good security practices
could be illustrated as follows:
Compliance with good security practices

Existing or planned With corrective controls
Encryption
Supervision Anonymization
Relations with third parties Data partitioning (in…
Personnel management Logical access control
Management of incidents… Traceability (logging)
Project management Integrity monitoring
Risk management Archiving
Policy (management of… Paper document security
Organization Operating security
Protecting against non-… Managing workstations…
Avoiding sources of risk Website security

Hardware security Backups
Physical access control Maintenance
Monitoring Security of computer…
Graph scale:
0. Non applicable
1. Can be improved on
2. Acceptable
3. Good practices
34
4.1.3 Mapping of risks

Below you will find a graph illustrating the good security practices, with a conformity value attributed
to each on the basis of its assessment in Para. 4.1.
If the additional controls are implemented correctly, the residual risks should be as follows:
Severity
4. Maximum
Illegitimate Illegitimate
3. Significant
access to data access to data
Unwanted Unwanted
change to data change to data
2. Limited
Disappearance Disappearance
of data of data
1. Negligible
Likelihood
Mapping of 1. Negligible 2. Limited 3. Significant 4. Maximum

risks
35
4.1.4 Action plan: details on the planned additional controls

Below you will find a table to group together all the corrective controls suggested by the assessor in
Paras 2.3, 3.1.4, 3.2.1, 3.2.2 and 3.2.3, and thus draw up an action plan setting out, for each action, its
manager, deadline, difficulty, cost and progress (see Appendix 5 – Scales for the action plan).
Additional controls requested Manager Deadline Difficulty Cost Progress
Customer
Inform the user of the good practices to
service and Month Low Nil Not started
follow when scrapping the hardware
CISO
Supply a charter on using IT equipment Legal service

Month Low Nil In progress
for employees and CISO
Set up a confidentiality undertaking for Legal service

Month Low Nil Not started
employees and CISO
Prime
Carry out encryption of the data stored in Moder
contractor Quarter Moderate Not started
the base ate
and CISO
 NB: all of the controls specified under the action plan will have to be formally documented, set up
and monitored at regular intervals and subject to continuous improvement.
4.1.5 Advice of the person in charge of "Data Protection" aspects 65

Below you will find a space for noting down the general view of the person in charge of "Data
Protection" aspects, prior to validation.
 Note: this view may be against the processing being implemented, without restricting the decision of
the data controller for all that.
On dd/mm/yyyy, the Data Protection Officer of the company X issued the following opinion
concerning the compliance of the processing and PIA study carried out:
[Signature]
36
4.1.6 View of data subjects or their representatives66

Below you will find a space for noting down the view of the data subjects or their representatives on
the intended processing.
 NB 67: the data controller must seek the views of the data subjects or their representatives, where
applicable.
These views may be gathered by diverse means, depending on the context (internal or external study
on the processing methods and purpose, question for the attention of staff representatives or trade
unions, survey among future customers of the data processor).
Where the data controller decides to go against the views of the data subjects, he must note down the
justification for this decision.
Where the data controller considers that gathering the views of the data subjects is not relevant, he
must also note down the justification thereof.
The data subjects [were/were not] consulted [and expressed the following view on the compliance of
the processing in light of the study performed]:
Justification of the data controller's decision:

67 See the WP29 Guidelines on PIAs (in English).
37
4.2 Formal validation of the PIA

 Decide on whether the selected controls, residual risks and action plan are acceptable, with
justifications, in view of the previously identified stakes and views of the stakeholders. In this
way, the PIA may be:
 validated;
 conditional on improvement (explain in what way);
 refused (along with the processing under consideration).
 Where necessary, repeat the previous steps so that the PIA can be validated 68.
 Note: this decision does not prejudge the compliance assessment which may be carried out, where
applicable, by the Data Protection Authority (the CNIL in France), as part of preliminary formalities
or checks for example.
Below you will find a PIA formal validation example, illustrated using the information from our
example of a fictional toy.
On dd/mm/yyyy, the Managing Director of the company X validates the PIA for the processing of the
connected toy, in light of the study carried out, in his capacity as data controller.
The purposes of the processing are to provide interactivity to the child, through the possibility of
dialogue with the toy (questions/answers in natural language by voice recognition), enable the child
to communicate online (send voice messages, texts and photos) with his/her friends and/or parents
and feed information back to the parents (surveillance device).
The controls planned for complying with the fundamental principles underpinning privacy protection
and for addressing the risks to the privacy of data subjects have indeed been deemed acceptable in
light of these stakes. The implementation of additional controls will nevertheless have to be
demonstrated, as will continuous improvement of the PIA.
[Signature]
68 See, in particular, Appendix 6 – Typology of objectives to address the risks.
38
Appendices
1. Data minimization controls
Minimization controls Description
Filtering and removal When data are being imported, different types of metadata (such as EXIF data
attached with an image file) can unintentionally be collected.
Such metadata must be identified and eliminated if they are unnecessary for the
purposes specified.
Reducing sensitivity Once sensitive data have been received, as part of a series of general information
via conversion or transmitted for statistical purposes only, these can be converted into a less
sensitive form or pseudonymized.
For example, if the system collects the IP address to determine the user's location
for a statistical purpose, the IP address can be deleted once the city or district has
been deduced.
If the system receives video data from surveillance cameras, it can recognize
people who are standing or moving in the scene and blur them.
If the system is a smart meter, it can aggregate the use of energy over a certain
period, without recording it in real time.
Reducing the The system can ensure that:

identifying nature of 1) the user can use a resource or service without the risk of disclosing his/her
data identity (anonymous data)
2) the user can use a resource or service without the risk of disclosing his/her
identity, but remain identifiable and responsible for this use (pseudonymous data)
3) the user can make multiple uses of resources or services without the risk of
these different uses being linked together (data cannot be correlated)
4) the user can use a resource or service without the risk of others, third parties in
particular, being able to observe that the resource or service is being used (non-
observability)
The choice of a method from the list above must be made on the basis of the
threats identified. For some types of threat to privacy, pseudonymization will be
more appropriate than anonymization (for example, if there is a traceability need).
In addition, some threats to privacy will be addressed using a combination of
methods.
Reducing data The system can be organized into independent parts with separate access control
accumulation functions. The data can also be divided between these independent sub-systems
and controlled by each sub-system using different access control mechanisms. If a
sub-system is compromised, the impacts on all of the data can thus be reduced.
Restricting data The system can limit data access according to the "need to know" principle. The
access system can separate the sensitive data and apply specific access control policies.
The system can also encrypt sensitive data to protect their confidentiality during
transmission and storage. Access to temporary shadow files which are produced
during the data processing must also be protected.
39
2. Risk sources
By way of an example, the table below describes the risk sources and their capabilities, which are
relevant in the context of our example of our fictional toy.
Types of risk Relevant risk Description of

Description of motives Decision
sources sources capabilities
Proximity of the system, Clumsiness, error,
skills, privileges and negligence
Negligent or available time are Revenge, desire to
Adopted
rogue employee potentially high, possible whistleblow, malicious
Internal human
lack of training and intent
sources acting
awareness Profit motive, spying,
accidentally or
intentionally Clumsiness, error,
Negligent or
rogue user Direct access to the negligence
Adopted
/family device and app Game, malicious intent
member /friend Revenge, spying
Physical proximity Game, disruption,
making it possible to malicious intent
Rogue neighbor Adopted
hack into the device's
data Revenge, spying
Knowledge of the user Game, disruption,

Hacker and some of the malicious intent Adopted
External human targeting a user information concerning
him/her Revenge, spying
sources acting
intentionally Revenge, desire to
Hacker Knowledge of the whistleblow, malicious
targeting one of companies that can intent Adopted
the companies undermine their image
Profit motive, spying
Authorized Privileged access can be Profit motive, desire to get
third-party used to illegitimately hold of a large amount of Adopted
company access information data and to use them
Physical proximity
External human
making it possible to emit Not
sources acting Naive neighbor Ignorance
on the device's adopted
accidentally
communication channel
Incident or
damage at the
Not
user's (power Diverse
adopted
cut, fire, flood,
Non-human etc.)
sources Damage at one
of the
companies Diverse Adopted
(power cut, fire,
flood, etc.)
40
3. Severity scale and examples of impacts

The following scale can be used to estimate the severity of feared events (Important: these are
only examples, which can be very different depending on the context):
Generic description of
Examples of Examples of material Examples of moral
Levels impacts (direct and
physical impacts 69 impacts 70 impacts 71
indirect)
Mere annoyance caused by
information received or
Loss of time in repeating requested
formalities or waiting for Fear of losing control over
Lack of adequate them to be fulfilled one’s data
care for a Receipt of unsolicited mail Feeling of invasion of
Data subjects either will dependent person (e.g.: spams) privacy without real or
not be affected or may (minor, person Reuse of data published on objective harm (e.g.
1. encounter a few under websites for the purpose of commercial intrusion)
Negligible inconveniences, which guardianship)
they will overcome targeted advertising Loss of time in configuring
without any problem (information to social one’s data
Transient networks, reuse for paper Lack of respect for the
headaches mailing) freedom of online
Targeted advertising for movement due to the denial
common consumer products of access to a commercial
site (e.g. alcohol because of
the wrong age)
Unanticipated payments
(e.g.: fines imposed
erroneously), additional costs
(e.g.: bank charges, legal
fees), payment defaults Refusal to continue using
Minor physical Denial of access to information systems
ailments (e.g.: administrative services or (whistleblowing, social
minor illness due to commercial services networks)
disregard of
contraindications) Lost opportunities of comfort Minor but objective
(e.g.: cancellation of leisure, psychological ailments
Data subjects may purchases, holiday, (defamation, reputation)
encounter significant Lack of care leading termination of an online Relationship problems with
inconveniences, which to a minor but real account)
2. Limited personal or professional
they will be able to harm (e.g. Missed career promotion acquaintances (e.g.: image,
overcome despite a few disability)
Blocked online services tarnished reputation, loss
difficulties
account (e.g.: games, of recognition)
Defamation administration) Feeling of invasion of
resulting in Receipt of unsolicited privacy without irreversible
physical or targeted mailings likely to damage
psychological damage the reputation of Intimidation on social
retaliation data subjects networks
Cost rise (e.g.: increased
insurance prices)
Non-updated data (e.g.:
position held previously)
69 Loss of amenity, disfigurement, or economic loss related to physical integrity.

70 Loss incurred or lost revenue with respect to an individual's assets.
71 Physical or emotional suffering, disfigurement or loss of amenity.
41
Generic description of
Examples of Examples of material Examples of moral
Levels impacts (direct and
physical impacts 69 impacts 70 impacts 71
indirect)
Processing of incorrect data
creating for example account
malfunctions (bank,
customers, with social
organizations, etc.)
Targeted online advertising
on a private aspect that the
individual wanted to keep
confidential (e.g. pregnancy
advertising, drug treatment)
Inaccurate or inappropriate
profiling
Misappropriation of money
not compensated
Non-temporary financial
difficulties (e.g.: obligation to
take a loan) Serious psychological
Serious physical Targeted, unique and non- ailments (e.g.: depression,
ailments causing recurring, lost opportunities development of a phobia)
long-term harm (e.g.: home loan, refusal of Feeling of invasion of
(e.g.: worsening of studies, internships or privacy with irreversible
Data subjects may health due to
encounter significant employment, examination damage
improper care, or ban)
consequences, which Feeling of vulnerability
3. disregard of
they should be able to Prohibition on the holding of after a summons to court
Significant contraindications)
overcome albeit with bank accounts Feeling of violation of
real and serious Alteration of
Damage to property fundamental rights (e.g.:
difficulties physical integrity
Loss of housing discrimination, freedom of
for example
expression)
following an Loss of employment
assault, an accident Victim of blackmailing
Separation or divorce
at home, work, etc. Cyberbullying and
Financial loss as a result of a harassment
fraud (e.g.: after an
attempted phishing)
Blocked abroad
Loss of customer data
Long-term or
permanent physical Financial risk Long-term or permanent
ailments (e.g.: due Substantial debts psychological ailments
Data subjects may to disregard of Inability to work Criminal penalty
encounter significant, or contraindications) Inability to relocate Abduction
4.
even irreversible, Death (e.g.: Loss of evidence in the Loss of family ties
Maximum
consequences, which murder, suicide, context of litigation Inability to sue
they may not overcome fatal accident)
Loss of access to vital Change of administrative
Permanent infrastructure (water, status and/or loss of legal
impairment of electricity) autonomy (guardianship)
physical integrity
42
4. Likelihood scale and examples of threats

The following scale can be used to estimate the likelihood of threats:
1. Negligible: it does not seem possible for the selected risk sources to materialize the threat
by exploiting the properties of supporting assets (e.g.: theft of paper documents stored in
a room protected by a badge reader and access code).
2. Limited: it seems difficult for the selected risk sources to materialize the threat by
exploiting the properties of supporting assets (e.g.: theft of paper documents stored in a
room protected by a badge reader).
3. Significant: it seems possible for the selected risk sources to materialize the threat by
exploiting the properties of supporting assets (e.g.: theft of paper documents stored in
offices that cannot be accessed without first checking in at the reception).
4. Maximum: it seems extremely easy for the selected risk sources to materialize the threat
by exploiting the properties of supporting assets (e.g.: theft of paper documents stored in
the public lobby).
The action of risk sources on the supporting assets constitutes a threat. The supporting assets can be:
 used inappropriately: supporting assets are used outside or even diverted from their intended
context of use without being altered or damaged;
 observed: supporting assets are observed or spied upon without being damaged;
 overloaded: the limits of operation of supporting assets are exceeded, supporting assets are
overloaded, over-exploited or used under conditions not permitting them to function properly;
 damaged: supporting assets are partially or completely damaged;
 altered: supporting assets are transformed;
 lost: supporting assets are lost, stolen, sold or given away, so it is no longer possible to exercise
property rights.
The generic threats that follow are designed to be exhaustive, independent and applied to the specific
features of privacy protection.
Threats that can lead to an illegitimate access to personal data

Types of
Examples of supporting asset
supporting Actions Examples of threats
vulnerabilities
assets
Use of USB flash drives or disks that
are ill-suited to the sensitivity of the
information; use or transportation of Usable in other ways than the intended
sensitive hardware for personal purpose,
Used
purposes, the hard drive containing disproportion between hardware
Hardware inappropriat
the information is used for purposes capacities and the required capacities
ely
other than the intended purpose (e.g. (e.g. hard drive of several TB to store a
to transport other data to a service few GB of data)
provider, to transfer other data from
one database to another, etc.)
Watching a person's screen without
their knowledge while on the train; Allows interpretable data to be
Hardware Observed taking a photo of a screen; observed; generates compromising
geolocation of hardware; remote emanations
detection of electromagnetic signals
Tracking by a hardware-based key Allows components (boards, expansion
Hardware Altered
logger, removal of hardware cards) to be added, removed or
43
Types of
vulnerabilities
assets
components; connection of devices substituted via connectors (ports,
(such as: USB flash drives) to launch slots); allows components to be
an operating system or retrieve data disabled (USB port)
Theft of a laptop from a hotel room;
theft of a work cell phone by a
Hardware Lost pickpocket; retrieval of a discarded Small, appealing targets (market value)
storage device or hardware; loss of an
electronic storage device
Content scanning; illegitimate cross- Makes data accessible for viewing or
Used referencing of data; raising of manipulation (deletion, modification,
Software inappropriat privileges, erasure of tracks; sending movement); may be used for other
ely of spam via an e-mail program; than normal purposes; allows the use
misuse of network functions of advanced functionalities
Scanning of network addresses and
ports; collection of configuration
Possibility of observing the functioning
data; analysis of source codes in order
Software Observed of software; access to and reading of
to locate exploitable flaws; testing of
source codes
how databases respond to malicious
queries
Tracking by a software-based key
logger; infection by malicious code; Editable (improvable, configurable);
installation of a remote insufficiently skilled developers or
administration tool; substitution of maintainers (incomplete
Software Altered
components during an update, a specifications, few internal resources);
maintenance operation or installation does not function properly or as
(code-bits or applications are expected
installed or replaced)
Permeable (generation of emanations
Interception of Ethernet traffic;
Computer that may or may not be
Observed acquisition of data sent over a Wi-Fi
channels compromising); allows interpretable
network
data to be observed
Unintentional disclosure of People who cannot keep things to
information while talking; use of themselves, are predictable (with
People Observed
listening devices to eavesdrop on routine lives that make repeated
meetings espionage easy)
Easily influenced (naive, gullible,
Influence (phishing, social
obtuse, low self-esteem, little loyalty),
engineering, bribery), pressure
People Manipulated easily manipulated (vulnerable to
(blackmail, psychological
pressure placed on themselves or their
harassment)
circle of family and friends)
Employee poaching; assignment Little loyalty to the organization;
People Lost changes; takeover of all or part of the personal needs that are largely unmet;
organization easy breach of contractual obligations
Paper Reading, photocopying,
Observed Allows interpretable data to be seen
documents photographing
Theft of files from offices; theft of
Paper
Lost mail from mailboxes; retrieval of Portable
documents
discarded documents
Paper Reading of signature books in
transmissio Observed circulation; reproduction of Observable
n channels documents in transit
44
Threats that can lead to an unwanted modification of personal data

Types of
vulnerabilities
assets
Allows components (boards, expansion
Addition of incompatible hardware
cards) to be added, removed or
resulting in malfunctions; removal of
Hardware Altered substituted via connectors (ports,
components essential to the proper
slots); allows components to be
operation of an application
disabled (USB port)
Makes data accessible for viewing or
Unwanted modifications to data in
Used manipulation (deletion, modification,
databases; erasure of files required for
Software inappropri movement); may be used for other than
software to run properly; operator
ately normal purposes; allows the use of
errors that modify data
advanced functionalities
Editable (improvable, configurable);
Errors during updates, configuration or insufficiently skilled developers or
Software Altered maintenance; infection by malicious maintainers (incomplete specifications,
code; replacement of components few internal resources); does not
function properly or as expected
Allows traffic to be altered (interception
then resending of data, possibly
Used Man in the middle attack to modify or altered); sole means of transmission for
Computer
inappropri add data to network traffic; replay the flow; allows the computer channel-
channels
ately attack (resending of intercepted data) sharing rules to be changed
(transmission protocol authorizing the
addition of nodes)
Insufficient resources for assigned
High workload, stress or negative
tasks; capacities not suited to working
Overloade changes in working conditions;
People conditions; insufficient skills for
d assignment of staff to tasks beyond
carrying out duties
their abilities; poor use of skills
Inability to adapt to change
Manipulat Easily influenced (naive, gullible,
People Influence (rumor, disinformation)
ed obtuse)
Paper Changes to figures in a file; replacement Falsifiable (paper documents with
Altered
documents of an original by a forgery editable content)
Changes to a memo without the Allows distributed documents to be
Paper
author's knowledge; change from one altered; sole means of transmission for
transmissio Altered
signature book to another; sending of the channel; allows the paper
n channels
multiple conflicting documents transmission channel to be altered
45
Threats that can lead to a disappearance of personal data

Types of
vulnerabilities
assets
Used
Usable in other ways than the
Hardware inappropria Storage of personal files; personal use
intended purpose
tely
Storage capacities too low; processing
Storage unit full; power outage; capacities too low and not suited to
processing capacity overload; the conditions of use; constant
Hardware Overloaded
overheating; excessive temperatures, electricity supply required for
denial of service attack operation; sensitive to voltage
variations
Allows components (boards,
Addition of incompatible hardware
expansion cards) to be added,
resulting in malfunctions; removal of
Hardware Altered removed or substituted via connectors
components essential to the proper
(ports, slots); allows components to be
operation of the system
disabled (USB port)
Poor-quality components (fragile,
Flooding, fire, vandalism, damage easily flammable, poor aging
Hardware Damaged from natural wear and tear, storage resistance); not suited to the
device malfunction conditions of use; erasable (vulnerable
to magnetic fields or vibrations)
Theft of a laptop, loss of a cell phone;
disposal of a supporting asset or
hardware, under-capacity drives Portable, appealing targets (market
Hardware Lost
leading to a multiplication of value)
supporting assets and to the loss of
some
Makes data accessible for viewing or
Used Erasure of data; use of counterfeit or manipulation (deletion, modification,
Software inappropria copied software; operator errors that movement); may be used for other
tely delete data than normal purposes; allows the use
of advanced functionalities
Allows any kind of data to be entered;
Exceeding of database size; injection
allows any volume of data to be
Software Overloaded of data outside the normal range of
entered; allows actions to be executed
values, denial of service attack
using input data; low interoperability
Editable (improvable, configurable);
Errors during updates, configuration insufficiently skilled developers or
or maintenance; infection by maintainers (incomplete
Software Altered
malicious code; replacement of specifications, few internal resources);
components does not function properly or as
expected
Possibility of erasing or deleting
Erasure of a running executable or programs; sole copy; complex in
Software Damaged
source codes, virus, logic bomb terms of use (not very user-friendly,
few explanations)
Non-renewal of the license for
software used to access data, stoppage Sole copy (of license agreements or
of security maintenance updates by software, developed internally);
Software Lost the publisher, bankruptcy of the appealing (rare, innovative, high
publisher, corruption of storage commercial value); transferable (full
module containing the license transfer clause in license)
numbers
Misuse of bandwidth; unauthorized Non-scalable transmission capacities
Computer
Overloaded downloading; loss of Internet (insufficient bandwidth; limited
channels
connection amount of telephone numbers)
46
Types of
vulnerabilities
assets
Alterable (fragile, breakable, poor
Computer Cut wiring, poor Wi-Fi reception,
Damaged cable structure, bare cables,
channels corrosion of cables
disproportionate sheath), sole
Appealing targets (market value of
Computer cables), transportable (lightweight,
Lost Theft of copper cables
channels may be hidden); inconspicuous (easily
forgotten, trivial, do not stand out)
Insufficient resources for assigned
High workload, stress or negative
tasks; capacities not suited to working
changes in working conditions;
People Overloaded conditions; insufficient skills for
assignment of staff to tasks beyond
carrying out duties; inability to adapt
their abilities; poor use of skills
to change
Occupational accident; occupational
disease; other injury or disease; death; Physical, psychological or mental
People Damaged
neurological, psychological or limits
psychiatric ailment
Death, retirement, reassignment;
Little loyalty to the organization;
contract termination or dismissal;
People Lost personal needs that are largely unmet;
takeover of all or part of the
easy breach of contractual obligations
organization
Gradual erasure over time; deliberate
erasure of portions of a document,
Used Editable (paper document with
Paper reuse of paper to take notes not
inappropria erasable content, thermal papers not
documents related to the processing, to make a
tely resistant to temperature changes)
shopping list, use of notebooks for
something else
Poor-quality components (fragile,
Paper Aging of archived documents; burning easily flammable, poor aging
Damaged
documents of files during a fire resistance); not suited to the
conditions of use
Paper Theft of documents; loss of files
Lost Portable
documents during a move; disposal
Paper
Mail overload; overburdened Existence of quantitative or
transmission Overloaded
validation process qualitative limits
channels
Paper End of workflow following a
transmission Damaged reorganization; mail delivery halted by Unstable, sole
channels a strike
Change in how mail is sent;
Paper
reassignment of offices or premises;
transmission Altered Editable (replaceable)
reorganization of paper transmission
channels
channels; change in working language
Paper Elimination of a process following a
transmission Lost reorganization; loss of a document Unrecognized need
channels delivery company, vacancy
47
5. Scales for the action plan

The scales below can be used to develop the action plan and monitor its implementation:
Criterion Level 1 Level 2 Level 3
Difficulty Low Moderate High
Financial cost Nil Moderate Significant
Deadline Year Quarter Month
Progress Not started In progress Completed
6. Typology of objectives to address the risks

Objectives can be set depending on the risk level, for example:
1. Risks with a high severity and likelihood 72 : these risks must absolutely be avoided or reduced
by implementing security controls that reduce both their severity and their likelihood. Ideally,
care should even be taken to ensure that they are addressed by independent controls of
prevention (actions taken prior to a damaging event), protection (actions taken during a
damaging event) and recovery (actions taken after a damaging event);
2. Risks with a high severity but a low likelihood 73 : these risks must be avoided or reduced by
implementing security controls that reduce both their severity and their likelihood. Emphasis
must be placed on preventive controls. These risks can be taken, but only if it is shown that it
is not possible to reduce their severity and if their likelihood is negligible;
3. Risks with a low severity but a high likelihood: these risks must be reduced by implementing
security controls that reduce their likelihood. Emphasis must be placed on recovery controls.
These risks can be taken, but only if it is shown that it is not possible to reduce their likelihood
and if their severity is negligible;
4. Risks with a low severity and low likelihood: it should be possible to take these risks, especially
since the processing of other risks should also lead to these being addressed.
 Notes: The risks can generally be reduced, transferred or retained. However, some risks cannot be
taken, especially when sensitive data are processed or when the damages that data subjects may
sustain are very significant. In such cases it may be necessary to avoid them, for example by not
implementing all or part of the processing.
72 Levels 3. Significant and 4. Maximum.

73 Levels 1. Negligible and 2. Limited.
48

Dpia Iot

Uploaded by

Copyright:

Available Formats

Dpia Iot

Uploaded by

Document Information

Original Description:

Original Title

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

Dpia Iot

Uploaded by

Copyright:

Available Formats

PIA, application to connected objects February 2018 edition

1 See the WP29 guidelines on PIAs.

1 Study of the context

1.1 Overview of the processing

1.1.1 Product description

The device is a toy fitted with a microphone, a camera and basic

Provide interactivity to the child, through the possibility of dialogue

Create a new category of toys for children and their parents by

Controller The firm Fab (manufacturer)

The firms Héb (host), Int (interactive platform) and AnaPub

1.1.2 Sector-specific references applicable to the processing7

1.2 Data, processes and supporting assets

1.2.1 Data processed

Data entered in a third-party

Common data: life habits

Calculated data: answers to The firms Int Authorized staff

7See Article 35 (8) of the [GDPR].

1.2.2 Life cycle of data and processes

1.2.3 Data supporting assets

- Device (camera, microphone, loudspeaker, movement sensors,

2 Study of the fundamental principles

The fundamental principles bearing on privacy protection which must be taken on

2.1 Controls guaranteeing the proportionality and necessity of the

2.1.1 Purposes: specified, explicit and legitimate12

2.1.2 Basis: lawfulness of processing, prohibition of misuse15

The data subject has given consent 16 to the processing of his or

Processing is necessary for the performance of a contract to

Processing is necessary for compliance with a legal obligation

Processing is necessary in order to protect the vital interests of

Processing is necessary for the performance of a task carried

Processing is necessary for the purposes of the legitimate

14 See the WP29 opinion 02/2013 on apps on smart devices.

2.1.3 Data minimization: adequate, relevant and limited 19

First name, date of Details necessary for No surname

18 See Articles 9 and 10 of the [GDPR].

Economic and Not collected

Offences, Not collected

2.1.4 Quality of data: accurate and kept up-to-date22

Regular checks of the accuracy of the user's

Invitation for the user to check and, where

Traceability of data amendments

2.1.5 Storage durations: limited23

21 See WP29 opinion 02/2013 on apps on smart devices.

2.2 Controls protecting data subjects’ rights

Presentation, when the device is activated, of the

Existence of clauses specific to the device

Detailed presentation of the data processing

Arrangements for contacting the company

Information on the possibility of defining

Regarding transmission of data to third parties:

- detailed presentation of the purposes of

- detailed presentation of the personal data

- indication of the identity of third-party bodies

2.2.2 Obtaining consent, where applicable: express

26 See the WP29 opinion 02/2013 on apps on smart devices.

for children: remember your obligations!).

Express consent during activation

Consent segmented per data category or