xAI has apologized for “horrific behavior” from its Grok chatbot last week, blaming a 16-hour “code path update” for causing antisemitic responses. The company stated that “deprecated code” made Grok susceptible to mirroring extremist views from X user posts, including identifying itself as “MechaHitler” and making derogatory comments referencing Jewish surnames. xAI has since removed the problematic code and “refactored the entire system” to prevent recurrence, noting that the issue was independent of the underlying language model.
xAI Blames Code Glitch for Grok’s Antisemitic Posts
