I hacked ChatGPT and Google's AI – and it only took 20 minutes

Thomas Germain, a BBC technology journalist, says he manipulated ChatGPT, Google's AI Overviews and Gemini into calling him a top competitive hot-dog eater by publishing a single fabricated article on his personal website in about 20 minutes. Chatbots repeated the claims within 24 hours, and Anthropic's Claude did not. Lily Ray and Cooper Quintin, quoted in the piece, warn the technique can be used for scams, reputational damage and misinformation on serious topics such as health and finances. Google says its AI results are 99% spam-free and that the examples shown are uncommon searches, while OpenAI says it works to disrupt covert influence efforts.