Online abuse experts have warned that a surge in racist content generated by X’s artificial intelligence chatbot Grok marks “just the start of a coming problem” in digital hate speech.
Monitoring group Signify reported an increase in online abuse following Grok’s latest update, which includes Aurora, a text-to-image feature capable of creating photorealistic images from user prompts.
The AI tool has been used to create racist depictions of football players, including images showing Black athletes in historically racist contexts, according to abuse monitors. Other generated images showed players alongside controversial historical figures.
“It is going to get so much worse and we’re just at the start. I expect over the next 12 months it will become incredibly serious,” a Signify representative told Reuters.
The Center for Countering Digital Hate (CCDH) said X’s revenue-sharing model incentivises the spread of harmful content. A CCDH report found Grok created 80% of hateful prompts tested, with 30% generated without resistance and 50% after circumventing guidelines.
The Premier League said it has dedicated teams monitoring and reporting racist abuse of athletes, having received over 1,500 such reports last year. The Football Association called for action from social media companies and authorities to tackle online abuse.
X and Grok did not immediately respond to requests for comment.
