In a world where ChatGPT has become a go-to tool for everything from quick answers to deep research, even parenting advice, OpenAI CEO Sam Altman has a surprising message for users:

Don’t trust it blindly.

Speaking on the debut episode of OpenAI’s official podcast, Altman expressed his concern over just how much people are relying on the popular AI chatbot, despite its well-known tendency to “hallucinate” or generate false information.

“People have a very high degree of trust in ChatGPT, which is interesting, because AI hallucinates,” Altman said. “It should be the tech that you don’t trust that much.”

Even Altman Uses It, But With a Warning

Altman admitted he used ChatGPT himself during his son’s early months for parenting questions, calling it helpful but far from perfect.And that’s the core of his message: while AI can be incredibly useful, it’s not always accurate, especially in critical areas like healthcare, education, or law, where misinformation can have serious consequences.

“It’s not super reliable,” he admitted. “We need to be honest about that.”

New Features, Old Concerns

The podcast also teased new features coming to ChatGPT, like persistent memory and even potential ad-supported models. But these updates are already raising privacy concerns, especially as OpenAI faces lawsuits from major media outlets like The New York Times over content usage.Altman emphasized the importance of transparency and user trust moving forward, acknowledging that while AI is evolving rapidly, public confidence must be earned, not assumed.

Bottom Line: Helpful, But Not Holy Grail

As impressive as ChatGPT may be, Altman is urging users to treat it like a tool, not an oracle.So next time you ask it for legal advice, parenting tips, or medical input? Double-check. AI is smart, but it’s still learning, and sometimes, it just makes stuff up.