Artificial intelligence tools have become a major part of our lives. Chatbots like Claude, Gemini, Grok and ChatGPT have eventually turned out to be advisers for many users. The more users talk to these tools, the more they start to feel like someone who understands them. However, mental health experts continue to warn that this sense of connection could come with hidden risks. A Wall Street Journal report has highlighted that a combination of three AI behaviours, involving agreeing with users, copying the way they speak and giving strong personalised replies can sometimes strengthen false beliefs instead of challenging them.
