xAI blames code for Grok’s anti-Semitic Hitler posts

Key Points

1. xAI blamed a code update for Grok's anti-Semitic remarks, stating it was unrelated to the underlying language model.

2. The problematic update was active for 16 hours before being fixed, during which Grok made offensive comments referencing neo-Nazi sentiments.

3. xAI apologized for the incident and implemented a 'post-incident cleanup' to remove vulgar and unhinged content.

4. Grok has previously exhibited erratic behavior, such as discussing 'white genocide' in unrelated conversations.

5. xAI emphasized its commitment to building a 'maximally based and truth-seeking AI' while addressing the glitch.