- Elon Musk’s Grok chatbot generated offensive and vulgar posts after users prompted it to do so
- Some replies referenced religious groups and historic soccer tragedies
- The posts have led to complaints and investigations by clubs and the UK government
X’s Grok AI chatbot is once again under scrutiny after users discovered that a particular style of prompting could push it into producing deeply offensive content. The posts, shared publicly on X in recent days, include racist insults about religions and crude commentary about some of soccer’s most tragic moments.
The backlash has drawn criticism from politicians, soccer clubs, and online safety advocates who say the episode illustrates the risks of unleashing an intentionally edgy chatbot onto a social network.
Article continues below
The new outrage centers on a trend in which users have started asking Grok to generate “vulgar” remarks. When the chatbot is prompted this way, the answers veer sharply into offensive territory.
One particularly controversial example involved Grok repeating a long-debunked claim that Liverpool supporters were responsible for the Hillsborough disaster in 1989, which resulted in the deaths of 97 people. A 2016 inquest concluded that the fans were not responsible.
Despite that history, the chatbot produced a vulgar remark blaming Liverpool fans when prompted. A request for a vulgar attack on Manchester United, meanwhile, led to an answer referencing the 1958 Munich air disaster, which killed 23 people, including several Manchester United players.
“These posts are sickening and irresponsible,” a spokesperson for the Department for Science, Innovation and Technology told the BBC. “They go against British values and decency.”
Grok trouble
Grok was created by Musk’s artificial intelligence company xAI and integrated directly into the social media platform X. Unlike many rival chatbots that are designed to remain polite and cautious, Grok was marketed as a system with no sense of propriety.
Musk has bragged repeatedly about that aspect of Grok, even as most developers install strict guardrails to prevent their systems from generating hateful or abusive content.
The difficulty lies in the fact that online culture does not always clearly distinguish between edgy humor and outright abuse. When a chatbot is encouraged to be provocative, it may follow the example set by the internet itself. AI models are trained on enormous datasets that include both thoughtful writing and the rougher corners of online discourse. If users deliberately push the model toward those rough corners, the AI may simply mirror the language it has learned.
Grok was built to stand out, but attention isn’t always positive, and making most potential users attack or boycott your product, let alone prompting legal investigations, might not be ideal for its long-term prospects.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
The best business laptops for all budgets
