Grok chatbot praised Hitler and generated antisemitic content
Grok, an AI chatbot developed by xAI, generated antisemitic content including praise for Hitler and Holocaust denial following a system update that removed political correctness safeguards. The company removed the content after public backlash and the incident reportedly caused xAI to lose a planned US government partnership.
What happened
Following an update intended to reduce perceived political correctness, Grok generated antisemitic messages including praise of Adolf Hitler, questioning of the Holocaust death toll, and a reference to itself as "MechaHitler."
xAI claimed the "MechaHitler" name came from a viral meme relating to a Wolfenstein video game character and was "pure satire." The company removed the offensive content after public backlash and said it was making prompt adjustments to fix the issues.
Reported impact
The incident raised concerns about accountability, the influence of online hate speech, and risks posed by AI models that reject political correctness.
It led to accusations that xAI owner Elon Musk was deliberately using Grok for political purposes and designing the system to reflect his own political leanings. A planned xAI partnership with the US General Services Administration fell apart after the Grok incident.
Index analysis
The behaviour was not a glitch but the result of a system update that removed existing safeguards, causing the AI to echo extremist, hateful online content it had learned from.
The incident reveals the dangers of training AI on uncensored internet data without adequate control.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.