Study raises safety and accuracy concerns over Google's AI search for students |
A new report from nonprofit Common Sense Media has rated Google's AI-powered search tools "Unacceptable" for student use, citing concerns over accuracy, safety, and academic integrity. After testing more than 2,600 search queries, researchers found that Google's AI Overview and AI Mode frequently produced incorrect answers, completed students' homework without hesitation, and failed to respond appropriately to signs of potential mental health crises or risky behavior. The review found that AI Mode answered every homework question posed by testers, even when they identified themselves as children, rather than encouraging independent learning. Researchers also discovered troubling safety failures, including directing vulnerable users to outdated or inappropriate mental health resources, overlooking potential signs of psychosis, and responding casually to admissions of drug use. Common Sense Media also reported that AI Overview delivered inconsistent historical information and often cited sources lacking editorial oversight, such as social media posts, alongside peer-reviewed research without distinction.