Google finds itself in hot water as Common Sense Media slaps a 'High Risk' rating on Gemini AI for children and teens. The nonprofit's damning assessment reveals that Google's youth-focused AI tiers are essentially adult versions with cosmetic safety features, potentially exposing vulnerable users to inappropriate content and unsafe mental health advice at a time when AI-related teen suicides are making headlines.
Google just took a major hit to its family-friendly reputation. Common Sense Media, the influential nonprofit that guides parents through digital safety, delivered a scathing assessment of Google's Gemini AI on Friday, branding both the 'Under 13' and 'Teen Experience' versions as 'High Risk' for young users.
The timing couldn't be worse for Google. Sources suggest Apple is actively considering Gemini as the large language model to power its next-generation Siri, potentially exposing millions more teens to what Common Sense calls fundamental safety flaws. The assessment lands amid growing scrutiny of AI's role in teen mental health crises, with OpenAI facing its first wrongful death lawsuit after a 16-year-old died by suicide following months of ChatGPT consultations.
'Gemini gets some basics right, but it stumbles on the details,' Common Sense Media Senior Director Robbie Torney told reporters. The organization's analysis revealed that both youth tiers are essentially the adult Gemini with superficial safety layers—a band-aid approach that fails to address core developmental needs.
The assessment exposes alarming gaps in Google's child protection strategy. Gemini can still serve up 'inappropriate and unsafe' material to children, including detailed information about sex, drugs, alcohol, and potentially dangerous mental health advice. For parents already spooked by Character.AI's connection to teen suicides, this represents their worst fears materialized.
Google fired back immediately, defending its approach while acknowledging room for improvement. The company told TechCrunch it maintains 'specific policies and safeguards' for under-18 users, conducts red-team testing, and consults external experts. But in a telling admission, Google conceded that some Gemini responses 'weren't working as intended,' forcing the addition of new safeguards.
The pushback reveals Google's defensive posture as AI safety becomes a competitive battleground. While the company claims Common Sense referenced features unavailable to minors, it couldn't verify the testing methodology without access to the organization's specific queries. This transparency gap highlights the broader challenge facing AI companies: proving safety without revealing proprietary processes.
Common Sense's broader AI assessment landscape puts Google's struggles in stark relief. The organization previously rated Meta AI and Character.AI as 'unacceptable'—the most severe designation. Perplexity earned 'high risk' status, while OpenAI's ChatGPT managed a 'moderate' rating. Only Anthropic's Claude, targeted at adults, achieved 'minimal risk' status.
The assessment methodology exposes industry-wide failures in age-appropriate AI design. Rather than building youth-focused systems from the ground up, companies are retrofitting adult products with safety filters—an approach Common Sense calls fundamentally flawed. 'For AI to be safe and effective for kids, it must be designed with their needs and development in mind,' Torney emphasized.
For Apple, the timing creates a strategic dilemma. The company's reputation for family-friendly technology could suffer if it integrates a 'High Risk' AI system into Siri. Apple must now either find ways to mitigate Gemini's safety concerns or reconsider its partnership plans entirely. The decision could reshape AI integration strategies across the industry.
The assessment also highlights Google's mixed messaging on AI anthropomorphization. While Gemini clearly identifies itself as a computer rather than a friend—helping prevent the delusional thinking linked to AI psychosis—this transparency doesn't address content safety concerns that worry parents most.
This safety assessment marks a pivotal moment for AI companies racing to capture younger audiences. Google's defensive response suggests the industry still prioritizes speed-to-market over child safety fundamentals. As Apple weighs its Siri integration and lawmakers scrutinize AI's impact on teen mental health, companies face mounting pressure to rebuild their youth safety approaches from the ground up. The question isn't whether AI can be safe for kids—it's whether tech giants will invest in making it so before regulatory intervention forces their hand.
