A study commissioned by the Ministry of Justice discovered almost 300,000 examples of hate speech posted on Finnish-language websites over a two month period.
Most of the hate speech was posted on smaller forums and message boards.
"Hate speech is not just a challenge for the social media giants. In this report, forums clearly stand out," said Miia Aaltonen from the ministry-backed Facts Against Hate project.
The report was based on analysis of roughly 12 million Finnish-language comments and articles published in September and October 2020.
During the two-month period of the study, nearly 300,000 incidences of hate speech were identified, 97 percent of which appeared on online messageboards.
Smaller forums hosting hate speech
The most significant platform was Ylilauta.org, a forum that claims to receive 1.5 million unique visitors every month, which accounted for 285,000 – or 96 percent – of the hate speech messages identified.
Yle contacted Ylilauta's administrators to ask about the website's hate speech policies. In an email response, Ylilauta pointed to the rules and guidelines published on the website.
Story continues after photo.
The rules state that Ylilauta does not allow users to post content that, among other things, is illegal or promotes illegal activity, incites, glorifies or organises violence, or that is inappropriate towards others.
The site said that Ylilauta is mainly moderated on a volunteer basis and that efforts were being made to eliminate illegal material as soon as possible.
After message boards, Twitter and Instagram posts were the next most common platforms for hate speech. Blogs, comments on news websites and Facebook posts altogether accounted for less than 0.02 percent of hate speech identified by the study.
A vocal minority
According to the analysis, 26 percent of all comments identified as hate speech by artificial intelligence featured the word "Muslim".
Mari-Sanna Paukkeri, CEO of Utopia Analytics which compiled the report, said AI's ability to recognise hate speech depended on the definition provided by human programmers.
"The researchers set out what counted as hate speech and then the material was fed to the artificial intelligence," she said.
According to the report, a significant finding is that a small number of writers appear to produce most hate speech. Ten individual users accounted for 11 percent of all hate speech identified.
"Everyone is able to influence the spread of hate speech by changing their own behaviour. A small group can get a lot of attention if posts are shared widely," said Facts Against Hate's Miia Aaltonen.