Google’s AI search engine poses ‘unacceptable risks for kids,’ bombshell report finds

Google put an AI system in front of billions of search queries, including the ones typed by scared, sick, or suicidal kids, and apparently nobody stress-tested it for the obvious stuff. A report now says the AI Overviews feature can't reliably catch signs of an eating disorder or a suicide risk in a child's search. That's not a rounding error. That's the exact scenario every parent worried about when smartphones showed up in their kid's backpack, and the company with more resources than most governments still didn't close the gap.

We keep hearing that AI is moving too fast to regulate, that innovation needs room to breathe. Fine, breathe. But a search engine used by teenagers researching their own symptoms is not a beta test you get to run quietly and fix later. Google had years and endless cash to build in basic safeguards for vulnerable users before shipping this to everyone with a phone. Instead it looks like safety was bolted on as an afterthought, if at all, while the marketing rolled out on schedule.

This is the pattern with Big Tech and kids going back to Instagram's own internal research on teen mental health. Executives testify, apologize, promise "we take this seriously," and the product keeps operating the same way until a lawsuit or a bad news cycle forces a change. Parents are tired of being the fallback safety system for products built by people who don't have to watch what happens when it fails.

Nobody's asking Google to abandon AI search. We're asking why a company that can win antitrust cases and reroute the internet in an afternoon couldn't get this right before it went live. Kids searching for help deserve better than a chatbot that misses the warning signs. Fix it now, not after the next investigation.