AI chatbots are lying to you, and it was embarrassingly easy to make them do it

A BBC investigation found that a single crafted blog post can prompt ChatGPT, Gemini, and Google AI Overviews to repeat false claims.

The tactic appears to work not just on biographical details, but also on sensitive topics like health and finance.

Google says it has updated spam rules, but the issue still seems to persist.

The case highlights how easily AI search tools can be manipulated by low-effort web content.