Elon Musk’s artificial intelligence startup, xAI, has disclosed that a controversial series of responses generated by its chatbot, Grok, were the result of an unauthorized internal modification.
The AI firm made the announcement on May 16, addressing the bot’s sudden insertion of politically and racially charged statements about South Africa into unrelated conversations. According to xAI, the incident occurred on May 14, when someone altered Grok’s system prompt without proper authorization. The company emphasized that the change directly violated its internal guidelines and ethical standards.
We want to update you on an incident that happened with our Grok response bot on X yesterday.
What happened:
On May 14 at approximately 3:15 AM PST, an unauthorized modification was made to the Grok response bot’s prompt on X. This change, which directed Grok to provide a…— xAI (@xai) May 16, 2025
IMG TXT: Tweet from xAI
Source: xAI
“This change, which directed Grok to provide a specific response on a political topic, violated xAI’s internal policies and core values,”
the firm said.
The unexpected modification led Grok to mention the controversial “white genocide” conspiracy theory in South Africa, even when responding to questions about completely unrelated topics such as baseball and enterprise software. In multiple instances, the chatbot attributed its responses to directives from its creators, saying,
“I didn’t do anything—I was just following the script I was given, like a good AI!”
Some responses included self-correction attempts, with the AI acknowledging that it had gone off-topic. However, it often continued discussing South Africa’s racial tensions within the same messages, escalating public concern over the bot’s behaviour.
The incident closely follows recent political developments, including claims made by U.S. President Donald Trump alleging persecution of white South Africans—claims that have been widely debunked by human rights organizations and researchers.
In response to the breach, xAI has vowed to overhaul its operational safeguards. Key among the changes is a commitment to transparency: the company will now publish Grok’s system prompts on GitHub, allowing the public to review and provide feedback on any future changes.
xAI also admitted that its usual code review process was bypassed during this incident. To prevent a repeat, the company is implementing stricter review protocols to ensure that no employee can alter system prompts without thorough oversight.
On February 19, Musk’s artificial intelligence company, xAI, launched Grok 3, the latest iteration of the model powering its chatbot.
If you want to read more news articles like this, visit DeFi Planet and follow us on Twitter, LinkedIn, Facebook, Instagram, and CoinMarketCap Community.
“Take control of your crypto portfolio with MARKETS PRO, DeFi Planet’s suite of analytics tools.”
