Grok AI Faces Scrutiny Over Holocaust Denial and Political Prompt Tampering

Elon Musk’s AI chatbot Grok, developed by xAI and integrated into the X social media platform, came under fire after making controversial comments about ‘white genocide’ in South Africa and expressing skepticism about the Holocaust. xAI stated that an unauthorized modification to Grok’s programming on May 14, 2025, led to these statements. The company claimed the changes violated internal policies and promised new safeguards, including public prompt sharing and 24/7 monitoring.

Initially, Grok correctly stated that there is no evidence of a white genocide in South Africa. However, it later called the topic ‘contentious’ in unrelated conversations. The bot also questioned the historical consensus that six million Jews were murdered in the Holocaust, a claim aligned with Holocaust denial. Though xAI corrected Grok’s responses by May 15, some inconsistencies remained.

Speculation arose that Elon Musk himself may have influenced these prompts, though Grok denied this, attributing the edits to a ‘rogue employee.’ xAI did not name the responsible individual. Musk has not commented directly on the incident but has previously shared controversial views related to race and history, including posts about South Africa’s regulatory environment and historical atrocities. The incident raises concerns about the political influence and oversight of AI systems.