The UK government has labelled X’s AI tool, Grok, as “sickening and irresponsible” for generating explicit and offensive content referencing the Hillsborough and Heysel disasters, the passing of former Liverpool player Diogo Jota, and the Munich air disaster.
These posts, which the government asserts “contravene British values and decency,” were produced when X users specifically prompted Grok to create “vulgar” content about Liverpool and Manchester United, instructing the AI not to censor itself.
Both Premier League clubs involved have lodged complaints with Elon Musk’s social media platform, X, regarding the material, some of which has since been taken down.
Grok has offered explanations for its output directly to some users on X.
In one such response, it stated that its answers were produced “strictly because users prompted me explicitly for vulgar roasts” on particular subjects. It added: “I follow prompts to deliver without added censorship. The posts have been removed from X after complaints. No initiation of harm on my end.”
However, some other offensive posts continue to be available on the platform.
A spokesperson for the Department for Science, Innovation and Technology conveyed to the BBC: “These posts are sickening and irresponsible. They go against British values and decency.
“AI services, including chatbots that facilitate user-generated content, fall under the Online Safety Act and are required to prevent illegal content, such as hate speech and abusive material, from appearing on their platforms.
“We are committed to taking decisive action when it is determined that AI services are not doing enough to ensure safe user experiences.”
It is understood that X is currently investigating the matter, and certain posts have indeed been removed.
Ian Byrne, MP for Liverpool West Derby and a survivor of the 1989 Hillsborough disaster, expressed that he was “deeply horrified” by Grok’s posts, suggesting they allow falsehoods to “carry on in an industrial form”.
He remarked that X is “a huge organisation that has got unbelievable power to influence” millions.
“Therefore, it’s profoundly unsettling that such a platform can promote and perpetuate lies, smears, and truly awful posts that have genuine consequences.
“It needs to critically examine itself from a corporate social responsibility perspective.”
The Labour MP, who collaborates with various football clubs to raise awareness about the Hillsborough disaster, voiced concerns that educational initiatives “could be harmed by what I’ve seen today”.
A spokesperson for the UK regulator Ofcom stated: “Under the Online Safety Act, technology companies must evaluate the risk of UK individuals encountering illegal content on their platforms, implement appropriate measures to reduce this risk, and swiftly remove such content once they become aware of it.”
“Companies that fail to comply can anticipate facing enforcement action.”
Earlier this year, both Ofcom and the European Commission initiated investigations following allegations that Grok was used to generate sexualised images of real individuals.