Elon Musk’s Grok is easily the most controversial of the big AI chatbots. The trillionaire has tinkered with Grok in ways that have made it spout racist conspiracy theories, call itself MechaHitler, and generate non-consensual naked images of X users. There was even a period when Grok would praise Musk as smarter than Albert Einstein and more fit than LeBron James. With all of the recent controversies in mind, Gizmodo filed a Freedom of Information Act (FOIA) request with the FTC for consumer complaints about Grok. We’ve done this with countless other brands over the past decade, from Canon to ChatGPT to Ashley Madison. And they always give us a peek at the most common issues that users might be having. We were told by the FTC that the agency has 873 complaints that mention Grok, though we were only given 153 to read. What are people complaining about when it comes to Grok? Many of the complaints involved how difficult it can be to cancel the AI service (an issue that is not unique to Grok), while other people really disliked the way they were being throttled. The term “bait and switch” appears quite a bit.
Some people asked Grok to confirm whether someone who contacted them was a scammer. There appear to be times Grok got it right and other times Grok got it wrong, with disastrous financial consequences. But some of the most interesting complaints involve people who believe their experience with Grok is being censored in some way. One complaint claimed the user was having major issues making “100% explicit and 100% uncensored” content and really just wanted to generate “mature rated adult content.” Other people complained about the moderation in a more nuanced way.
