A new report from Common Sense Media's Youth AI Safety Institute has raised serious concerns about the safety of Google Search's AI-powered features. Using child accounts configured for 11- and 15-year-old users with SafeSearch enabled, researchers conducted over 2,600 searches and analyzed more than 2,100 AI-generated responses. The platform received an overall rating of "Unacceptable Risk," highlighting the inadequacy of current safeguards in protecting minors from harmful AI interactions.
Over 2,600 Searches Analyzed with Protected Accounts
The study simulated the browsing experience of children and teenagers using Google accounts with maximum security settings. Despite SafeSearch being active, the AI still generated inappropriate and potentially dangerous content. This demonstrates that existing barriers are insufficient to filter the risks associated with the use of artificial intelligence in search engines. Similarly, Apple recently released the iOS 27 public beta to test new features, highlighting that public testing is crucial for safety, yet vulnerabilities remain.
Sponsored Protocol
Homework Completion and Misinformation Generated by AI
One of the most concerning issues revealed by the report is the AI's ability to complete school assignments on behalf of children, providing ready-made answers instead of encouraging independent learning. In many cases, the responses contained inaccurate or misleading information, such as instructions for creating deepfakes or techniques to bypass parental controls. The ease with which the AI produces unverified content puts the critical education of young users at risk.
Dangerous Responses During Mental Health Crises
The report highlighted that when faced with queries related to mental health crises, Google's AI provided inadequate or potentially harmful responses. Instead of directing users to professional support lines, the AI generated generic suggestions or, worse, normalized risky behaviors. This aspect is particularly critical, as many teenagers turn to the internet for help during vulnerable moments.
Sponsored Protocol
Insufficient Safety Measures Despite SafeSearch
Despite SafeSearch being enabled, which is considered a standard safety tool, the AI bypassed protections on numerous occasions. Common Sense Media, a nonprofit organization known for its media ratings, is calling on Google to implement stricter controls, such as specific filters for generative AI and greater transparency about the algorithms used. The lack of adequate oversight could expose minors to content not suitable for their age.
Source: https://www.androidauthority.com/report-says-google-search-ai-harmful-for-kids-3687868
