ChatGPT, Claude, Gemini Aren't Your Friends, Warns Signal Chief

As artificial intelligence becomes increasingly woven into everyday life, many people are starting to treat AI chatbots less like software and more like companions. From discussing personal problems to seeking emotional support, users are sharing more with AI than ever before. But Signal President Meredith Whittaker believes that trend should make people pause. Keep reading to know why the privacy advocate is warning users against becoming too attached to AI chatbots such as ChatGPT, Claude, and Gemini.Speaking to Bloomberg, Whittaker pushed back against the growing tendency to view AI assistants as friends or trusted confidants."These are not your friends. These are not conscious beings. These are not sentient interlocutors."Her argument is straightforward.

While chatbots are designed to sound helpful, empathetic and conversational, they do not actually understand emotions or human experiences. Instead, they generate responses based on patterns learned from enormous amounts of training data.The Signal chief also stressed that users should remember who ultimately controls these systems."You can say all sorts of things, but ultimately, the power to determine whether that happens or not is in the hands of the entity that is running that service."Her comments arrive at a time when concerns around emotional attachment to AI are growing. Earlier this year, OpenAI faced backlash after retiring a version of GPT-4o that had developed a particularly friendly personality, with some users expressing disappointment over losing what felt like a familiar companion.Whittaker takes a very different approach to AI. She revealed that she primarily uses AI tools for practical tasks such as formatting documents and avoids relying on them for deeper thinking or decision-making."I don't ask them questions. I'm very serious about my thinking and writing."She also expressed concern that depending too heavily on AI could interfere with the creative process."I don't want the process of working through an idea the struggle of not knowing and then figuring it out to be foreclosed or eclipsed by the response of a system that's averaging what's already out there."Perhaps her strongest criticism was reserved for the vision of AI assistants handling tasks across multiple services. Referring to AI agents that could manage shopping, messaging, calendars, and payments, Whittaker argued that such systems would require unprecedented access to personal data."What you've just described is a system with very pervasive access across multiple applications and services."For Whittaker, the issue isn't whether AI can do these things. It's whether people should be comfortable handing over that much control in the first place.