Steam accused of hosting over 1.8 million instances of extremist and hateful content without proper moderation

The Anti-Defamation League (ADL) has conducted research into Steam. The organization found over 1.8 million instances of extremist or hateful content, stating that the platform needs to tackle these issues.

The ADL’s Center on Extremism analyzed 1.2 billion unique pieces of content on Steam using its proprietary machine learning tool called HateVision. As the organization explained in its study, this includes over 458 million user profiles, 152 million avatar images, and 610 million comments.

Below are the key takeaways from the ADL analysis:

• The organization identified 1.83 million instances of extremist or hateful content, including antisemitic symbols, swastika and other Nazi imagery, as well as images of terrorist organizations;
• 1.5 million users and over 73k groups used at least one potentially extremist or hateful symbol, copypasta, or keyword;
• There are 1.18 million hateful copypastas (in multi-lined text fields, including ASCII art), 51% of which were variations of swastikas;
• Over 827k user and group avatars contain extremist or hateful symbols, with thousands of profiles glorifying “violent extremists, like white supremacist mass shooters.”

In its report, the ADL also links extremist content found on Steam to real-life violence. The organization specifically cited an attack in August in which a teenage boy stabbed five people in Turkey. Local media claimed that he was “influenced by war games,” with the ADL also identifying his Steam profile containing various extremist content.

While the organization didn’t provide direct evidence that extremists are using Valve’s platform to organize and plan crimes, it told Bloomberg that allowing such content “increases the likelihood that someone will travel all the way down the rabbit hole.”

According to the ADL, “major gaming companies are selling their wares on a platform that is not addressing users’ support of extremists and allowing the proliferation of hate.”

The research cites flaws in Steam’s content and moderation guidelines that allow users to bypass the restrictions by removing keywords from the block list and using extremist symbols in their profile descriptions.

The ADL believes that Valve “needs to make significant changes to their approach to platform governance,” including following these recommendations:

• Adopt policies prohibiting forms of extremism and the expression of hateful ideologies;
• Clearly define consequences for violating the rules;
• Audit its moderation practices to close loopholes and consult with civil society groups and researches.

“In rare notable cases, Steam has selectively removed extremist content, largely based around extremist groups publicized in reporting or in response to governmental pressure,” the ADL stated. “However, this has been largely ad hoc, with Valve failing to systematically address the issue of extremism and hate on the platform.”

As Extremism & Gaming Research Network co-founder Galen Lamphere-Englund told Bloomberg, Steam moderation “tends to focus on things like fraud or money laundering” rather than extremist and hateful content.

The ADL’s research is not the first time that various non-governmental organizations or US officials have accused Steam of hosting such content. In 2022, Senator Maggie Hassan called on Valve founder Gabe Newell to take steps to prevent harmful content on the platform, also drawing parallels “between online comments and violence in the offline world.”

Similar concerns have been raised regarding other game companies. Last month, short-selling firm Hindenburg Research published a major report on Roblox, accusing the company of bloating its DAU and engagement metrics, describing the platform as “X-rated pedophile hellscape, exposing children to grooming, pornography, violent content and extremely abusive speech.”

This led to Roblox updating its child safety policies and introducing new accounts with parent privileges, among other measures.