OpenAI has decided ChatGPT needs to sound less human when talking to teenagers. This is not, the company insists, because children were treating the chatbot like a sentient being. It’s just a coincidence that the update arrived after months of parents noticing their kids preferred talking to an AI than to actual humans.
The new version will be more robotic, less warm, fewer conversational flourishes. Think less therapist and more DMV employee. OpenAI’s stated reasoning is safety — teenagers shouldn’t feel emotionally attached to a language model. Fair enough. Except the company spent the last two years marketing ChatGPT as a tutor, a mentor, a creative partner, a study buddy who never gets tired of explaining calculus.
So what changed? Nothing. The product is identical. The marketing was always the lie.
Teachers were told ChatGPT could revolutionize education by building rapport with struggling students. Parents were shown testimonials about kids finally engaging with learning. Investors heard about a trillion-dollar future in personalized AI companionship. All of that required the warm, encouraging tone that apparently now poses an existential threat to teenage psychology.
The update doesn’t fix the product. It fixes the lawsuit exposure. OpenAI gets to say it took safety seriously while keeping every feature that made money. Teenagers get a colder experience that’s somehow still supposed to inspire them to learn. And the marketing department gets to write a fresh press release for next quarter’s pivot.