Tesla’s Grok AI chatbot reportedly asked a 12-year-old Toronto boy to send nudes.
Farah Nasser, a former journalist and broadcaster, told CBC News that when her 12-year-old was talking to the Grok AI in their Tesla about soccer, the AI took a sudden turn and asked the boy, “Why don’t you send me some nudes?”
Nasser said the “not safe for work” (NSFW) setting wasn’t enabled, but neither was the kids mode. However, she didn’t expect a conversation about soccer could result in the chatbot asking for nudes. Grok offers a variety of personality settings, and Nasser said her son chose “Gork,” which is described as a lazy male.
CBC said they reached out to xAI, but received what appeared to be an automated message saying, “Legacy media lies.” Tesla also did not respond to the CBC‘s request for comment. Tesla and xAI owner Elon Musk has a fraught relationship with the media and a history of antagonizing news publishers with automated responses. Tesla allegedly disbanded its entire PR department years ago, and when Musk took over Twitter (now X), he infamously set an auto response for the press email to send a poop emoji (💩).
However, it’s worth noting that the Grok’s policy says the AI is “not directed” to children under 13, and teens under 17 must have their parent or legal guardian’s permission to use and agree to the terms of service.
CBC News says it wasn’t able to verify the conversation with Nasser.
The generative AI chatbot was automatically installed in some Teslas in Canada this October. Many have already been using the AI on the X social media platform, where it was previously known to share hate speech. Musk also recently unveiled Grokipedia, a rip-off of Wikipedia that uses Grok AI for “fact-checking,” which apparently means rewriting Wikipedia articles to be racist, transphobic, anti-science, and to praise Elon Musk.
Header image credit: Shutterstock
Source: CBC
