A new CyberWell report released Monday says AI-generated anti-Semitic content has proliferated across social media during a 13-month period marked by the rapid proliferation of AI image, video, and audio generators.
The report examined 300 cases of verified AI-generated anti-Semitic content, reaching more than 30 million views and generating more than 2.8 million engagements on major social media platforms, CyberWell said.
4 View gallery
AI-generated anti-Semitic social media posts
CyberWell said its first-of-its-kind investigation examined the intersection of generative AI abuse, social media governance, and social media usage. It highlights how repeated anti-Semitic abuse of AI tools affects large audiences on major platforms.
The report analyzed hundreds of AI-generated anti-Semitic posts published between January 2025 and February 2026. CyberWell said widely available tools such as OpenAI’s Sora, Google’s Veo, X’s Grok, and Suno were used to mass-produce convincingly anti-Semitic content.
AI tools blended fabricated visuals, audio and narratives with real-world footage, making anti-Semitic content more believable, more appealing and harder to detect, the report said.
According to CyberWell, much of the content that mocks or denies the Holocaust appears to target children and was found alongside content aimed at grooming or sexualizing minors in social media feeds. OpenAI ultimately shut down the tool following widespread abuse of Sora to spread problematic content, highlighting the difficulty of preventing abuse of advanced generative AI systems.
4 View gallery
According to the report, 79% of the anti-Semitic content analyzed appeared on video-based platforms such as TikTok, Instagram, and YouTube. Short-form videos dominate, often targeting younger audiences through gaming culture, parody content, and viral audio trends.
CyberWell said AI-generated music and gaming formats associated with platforms like Roblox and Minecraft have been repeatedly used to embed anti-Semitic narratives in online spaces aimed at young people.
AI-generated anti-Semitic content existed before Hamas’s attack on Israel on October 7, 2023, but has since gained more attention and amplified on social media, Cyberwell said. Due to the increased visibility of such content after the genocide, CyberWell began monitoring content as a regular part of its investigative methodology.
The report said there was a major turning point during the conflict involving Iran, Israel and the United States in June 2025, Cyberwell said, noting that it detected a sharp spike in the cross-platform spread of AI-generated anti-Semitic content.
4 View gallery
“Artificial intelligence has fundamentally changed the scale and speed at which anti-Semitism is created and distributed online,” said Tal Ór Cohen Montemayor, CEO and founder of Cyberwell. “Generative AI allows bad actors to industrialize hate and create high-impact content that can reach millions of people. Enforcement often occurs only after hate has already been widely amplified.”
The pattern of involvement was particularly alarming, the report said. Posts glorifying or calling for violence against Jews accounted for 33% of content and 41% of total engagement, indicating that the most extreme content spread the fastest and reached the widest audience.
Based on its previous State of Online Antisemitism 2025 report, CyberWell said that violent anti-Semitic content is twice as likely to appear in AI-generated content than in user-generated anti-Semitic content.
The report also highlighted inconsistent enforcement by platforms. TikTok had the largest share of content at 36% and also had the highest enforcement rate at 88%. Instagram accounted for the largest share of engagement at 65%, disproportionate to post shares at 25%.
4 View gallery
Tal-o-Coen Montemayor
(Photo: Hadar Bader)
Instagram’s parent companies, TikTok and Meta, recorded significantly higher than average removal rates for anti-Semitic posts, likely in part because they have clearer policy frameworks for dealing with AI-generated content, CyberWell said. In contrast, YouTube and X have lower takedown rates of 28% and 20%, respectively, and do not maintain similarly clear policy terms, the report said.
CyberWell said these fees reflect final regulations after content is reported to the platform, but enforcement is often delayed. AI-generated content often remains online long enough to accumulate hundreds of thousands or millions of views before being removed, the report said.
“Platforms need to go beyond disclosure and invest in systems that identify harmful narratives at scale, including those embedded in audio, visual, and coded formats that evade traditional detection,” Cohen-Montemayor said.
“Generative AI is a powerful technology, but it is also being weaponized on a large scale,” she said. “By strengthening automated detection, investing in competent and transparent human controls, auditing training data, and engaging external expert stakeholders, platforms and AI developers can address complex and rapidly evolving forms of online hate through continued collaboration with technology companies, policymakers, and expert partners.”
CyberWell is an independent, technology-based nonprofit organization working around the world to combat the spread of anti-Semitism online. The company’s AI technology monitors social media in English and Arabic for posts that promote anti-Semitism, Holocaust denial, and promote violence against Jews and their allies, based on the International Holocaust Remembrance Alliance’s (IHRA) Working Definition of Anti-Semitism. The company’s analysts will review and report this content to platform moderators, as well as index all verified posts in the first-ever open database of anti-Semitic social media posts.
