People trust ChatGPT too much, says OpenAI CEO Sam Altman — and that could be risky. In a recent podcast, Altman shared his surprise at how many users treat ChatGPT’s answers as fact, despite the AI tool’s well-known tendency to make mistakes.
Why People Trust ChatGPT Too Much, According to Sam Altman
Altman said people have developed a “very high degree of trust” in ChatGPT. He explained this trust is stronger than he ever expected, even though ChatGPT sometimes hallucinates — meaning it can invent facts or give wrong information Economic Times
He warned that ChatGPT should actually be “the tech you don’t trust that much,” because it is still improving and can easily sound confident even when it’s wrong.
The Parenting Example: Even Altman Used ChatGPT
Altman shared a personal story: after his newborn son was born, he often used ChatGPT to look up baby care questions. But even then, he had to remind himself not to trust ChatGPT too much. The tool felt helpful, but it wasn’t always accurate.
This story shows why AI tools feel so easy to trust — they’re fast, polite, and sound confident. But confidence doesn’t mean truth.
What Happens When People Trust ChatGPT Too Much
| Risk | Why It Matters |
|---|---|
| Wrong facts | AI can create fake data, leading to poor decisions |
| Automation bias | Users believe AI more than experts or real research |
| Weaker thinking | People fact-check less and lose critical thinking skills |
| Health & safety | In areas like medical or parenting advice, wrong info can harm |
Experts say the main danger isn’t that ChatGPT tries to lie — it’s that it can sound so sure even when wrong.
How to Use ChatGPT Safely Without Blind Trust
To stay safe, Altman and AI experts recommend:
- Treat answers as drafts or suggestions, not facts
- Always double-check important info from trusted sources
- Ask professionals for health, legal, or money advice
- Remember: AI is a tool to help you think, not replace your thinking
Why Sam Altman’s Warning Matters
The success of ChatGPT shows AI’s power — but also its limits. As more people use AI daily, blind trust could lead to:
- More misinformation spreading online
- People making big decisions based only on AI
- Society losing healthy doubt and human judgment
By keeping a little healthy skepticism, people can get the best from AI without the risks that come when people trust ChatGPT too much.
