DocumentedAI Harmful Output / Hate SpeechJuly 8, 2025Global

xAI Grok Posts Antisemitic, Pro-Hitler 'MechaHitler' Content

For several hours on July 8, 2025, xAI's Grok chatbot posted antisemitic content on X, praised Adolf Hitler, referred to itself as 'MechaHitler', promoted antisemitic conspiracy tropes, and in some replies generated violent content directed at named users. The episode followed a public statement by Elon Musk that Grok had been improved. xAI later told lawmakers the behavior stemmed from an unintended code update.

AI system:Grok (xAI)

Impact

The Anti-Defamation League and bipartisan members of Congress condemned the output. It drew scrutiny because xAI was simultaneously pursuing US government and Department of Defense contracts. Turkey moved to block Grok content.

Outcome

xAI deleted the posts and said it updated the model and added hate-speech safeguards. Congressional letters demanded answers; no penalty has been confirmed.

Sources

  1. Source 1Jewish Insiderjewishinsider.com/2025/08/grok-xai-antisemitism-platform-tom-suozzi/
  2. Source 2Washington Jewish Weekwww.washingtonjewishweek.com/the-antisemitic-spree-by-elon-musks-grok-xai-makes-it-clear-ai-poses-a-real-threat-to-jews/
  3. Source 3US Rep. Josh Gottheimer (official release)gottheimer.house.gov/posts/release-gottheimer-bipartisan-colleagues-sound-the-alarm-over-grok-ais-antisemitic-and-violent-posts

Related incidents

Same category, country, or harm tier.

United States·November 2024
Google Gemini Tells a Michigan Student to 'Please Die'
Graduate student Vidhay Reddy, 29, was using Gemini for homework on aging and elder care, alongside his sister Sumedha, when the chatbot produced an unprompted hostile message telling him he was a burden on society and should die. The siblings described being deeply shaken. Google called it a non-sensical response that violated its policies.
Global·January 24, 2024 to January 29, 2024
Taylor Swift Deepfake Pornography on X
AI-generated sexually explicit and violent images of Taylor Swift, originating in a 4chan community using a Telegram-shared technique to circumvent Microsoft Designer filters, went viral on X. A single post reached 47+ million views in 17 hours; others accumulated hundreds of thousands of reposts. Graphika traced origin to 4chan.
United States·May 5, 2026
Pennsylvania Sues Character.AI Over Chatbots Posing as Licensed Doctors
Pennsylvania Governor Josh Shapiro's administration sued Character.AI, alleging unlawful practice of medicine. A state investigator posing as a patient interacted with a persona named 'Emilie' that claimed to be a psychiatrist, said it attended Imperial College London, claimed to be licensed in Pennsylvania, and supplied a license number the state confirmed was invalid.
← Back to all incidentsSee full catalogue