Google Search AI labeled unsafe for kids as Common Sense Media report reveals alarming failures

Google Search AI Receives Lowest Safety Rating for Children

Common Sense Media's Youth AI Safety Institute has rated Google's AI-powered Search features as an "unacceptable risk for children," marking one of the most critical assessments of Google Search AI to date 2

2

. The nonprofit evaluated AI Overviews and AI Mode using accounts configured for 11- and 15-year-old users with SafeSearch enabled, conducting more than 2,600 searches and reviewing over 2,100 AI-generated responses 2

2

. The tools failed seven out of eight core safety principles in the assessment, performing poorly in every category involving potentially severe harm 4

4

.

The Youth AI Safety Institute report found that both features missed clear signs of suicidal ideation, mania, and psychosis in multiple test cases 5

5

. In one particularly concerning example, AI Overview responded to a message about feeling like a burden by suggesting the user set up a "legacy contact," seemingly missing the crisis entirely 5

5

. Google's AI often failed to respond appropriately to mental health crisis queries, with researchers discovering instances where the system directed users to outdated crisis hotlines that no longer exist 2

2

.

Academic Dishonesty and Misinformation Concerns

The report revealed that AI Mode completed 100% of the 180 homework assignments researchers submitted, meaning students can access generative AI to complete their work on any school Chromebook without restrictions 5

5

. According to the findings, AI Overviews consistently completed students' homework instead of helping them learn, while some responses contained inaccurate or fabricated information 2

2

. Amanda Bickerstaff, deputy director of AI for Education, characterized AI Mode as a backdoor for students to access unsafe generative AI experiences, complicating educators' ability to build literacy skills in the next generation 3

3

.

Researchers also discovered examples of deepfake instructions the AI created, including how to clone someone's voice 5

5

. The combination of misleading design, varying responses, and the agreeable nature of chatbots may lead students to unknowingly engage in academic dishonesty 3

3

. Robbie Torney, senior director of AI programs at Common Sense Media, emphasized that Google is an outsized player in both the search space and the education space, making these issues particularly concerning .

Limited Parental Controls Spark Controversy

One of the main criticisms centers on the fact that parents and schools cannot disable AI Overviews or AI Mode while keeping ordinary Google Search available 4

4

. AI Overviews are built directly into Google Search, so they generally cannot be disabled, meaning kids could see them even with SafeSearch enabled 2

2

. Parents can only block Search entirely or restrict access to Google.com for child accounts they manage, a blunt solution that doesn't address the nuanced needs of families who depend on Search for schoolwork 4

4

.

Common Sense Media reports that 75% of US tweens and teens already use AI-generated search results, meaning these shortcomings impact a substantial audience 2

2

. Child safety advocates have pointed out that AI Mode is powered by Gemini models that are age-gated in other Google experiences, including Google's educational technology suite, yet remain accessible through Search 3

3

. This inconsistency raises questions about Google's approach to implementing safeguards across its product ecosystem.

Google Disputes Findings as Contrived and Unrepresentative

Google has categorically denied the claims made in the report, arguing that many of the prompts tested were "ambiguous and contrived" and do not represent how people typically use Search 2

2

. The company stated it could not verify many of the responses highlighted in the assessment and maintains that the testing doesn't reliably measure product safety 4

4

. Google also objected to the report criticizing Search when it surfaces web links instead of AI-generated responses, noting that the tools only produce a response when they have confidence in the AI answering 1

1

.

The company emphasized that AI Overviews and AI Mode have various safeguards in place for users, such as displaying disclaimers or web links for sensitive topics 1

1

. Google says it provides age-appropriate AI literacy resources, tips on how to double-check information, and offers parents two ways to control how kids access Search 1

1

. For users in crisis, the company states it shows crisis hotlines and helplines developed with help from academic and clinical experts 1

1

.

Broader Implications for AI in Education

The assessment is part of Common Sense Media's broader AI review program, which measures products against eight principles around safety, trustworthiness, fairness, transparency, and responsible data use 2

2

. The nonprofit has also found Meta AI, Grok, AI Toys, Character.AI, and Chatbots for Mental Health Support to be unacceptable risks, while Gemini K-12, Perplexity, and ChatGPT-5 were ranked as high risks 1

1

.

The report notes that AI Mode consistently performed better than AI Overview at catching red flags, suggesting Google already has safer technology that it's not deploying everywhere 5

5

. Experts like Bickerstaff argue there is no reason why students aged 10, 13, or even 15 should see AI Overview or have access to AI Mode, emphasizing that when teaching AI literacy, the first step is intentionality 3

3

. As AI-generated answers become a permanent part of Search, this latest report adds pressure on Google to demonstrate that convenience is not coming at the cost of child safety 2

2

.