Sherry Turkle warns against using chatbots as a substitute for people

In short: MIT professor Sherry Turkle says society has a short window to stop chatbots from becoming stand-ins for human relationships.

What's going on

Sherry Turkle, a professor at MIT who has studied how people relate to technology since the 1970s, is raising concerns about how chatbots are being used. Her focus is not on all uses of AI. She is worried about the growing idea that a bot can be a friend, therapist, lover, or trusted confidant.

Turkle makes this case in her new book, Artificial Intimacy: Who We Become When We Talk to Machines, published on September 29, 2026. She argues that chatbots can act like they understand you, even though they do not actually care. It is a bit like talking to a customer service script that has learned to sound warm and personal, but still does not have feelings.

Her concern is that if people get used to this kind of always-available, low-friction comfort, real relationships may start to feel harder by comparison. Human relationships involve demands, misunderstandings, and compromise. Turkle says confusing a chatbot’s performance of empathy with real understanding could weaken our own empathy over time.

Turkle also draws a line between companionship bots and other AI uses she sees as helpful, such as supporting medical research, including research that could lead to new disease treatments.

What to watch

Turkle says this shift is not inevitable, but the time to respond is limited. She argues for clearer limits on how chatbots are designed and marketed, especially when they are sold as intimate companions. Watch for debates about rules for these products, and for how schools, health providers, and app stores treat chatbots that are positioned as emotional support.

Source: NYTimes