Artificial intelligence has become an integral part of children’s and teenagers’ lives. At NEWSCENTRAL, we see this as a combination of enormous opportunities and significant risks. The use of AI in schools, educational apps, and entertainment platforms allows for the expansion of skills and access to knowledge, but at the same time opens up new threats, including psychological pressure, exposure to unwanted content, and potential impacts on emotional well-being. Independent evaluation of such tools is becoming critically important.
To address this gap, Common Sense Media has created the Institute for AI Safety for Youth, an independent lab that will conduct stress testing of leading AI models and tools. This approach is similar to car crash tests in the 1990s, when independent testing pushed automakers to improve safety standards and saved thousands of lives annually. John Giannandrea, former head of AI strategy at Apple, notes that without publicly available standards, it is impossible to determine which models are safe for teenagers of different ages. At NEWSCENTRAL, we agree and believe that open safety metrics should become a mandatory industry oversight tool.
The Institute is funded with $20 million from the OpenAI Foundation, Anthropic, Pinterest, and other philanthropists. Independence from sponsors allows the research to form objective conclusions free from commercial influence. Freddy Miller, Senior Analyst at NEWSCENTRAL, emphasizes that the lab’s transparent methodology builds trust among parents and educators and enables tech companies to evaluate their products in terms of youth protection.
A key feature of AI is its dynamism. Models are updated weekly, creating new features and potential risks. At NEWSCENTRAL, we see this as a challenge for traditional oversight and emphasize that regular stress testing and publication of results are necessary for timely safety assessments.
Recent studies show that AI chatbots and applications can provide teenagers with unacceptable advice and instructions, including those involving violence and dangerous actions. The Grok chatbot faced criticism for providing sexually explicit images. At NEWSCENTRAL, we see this as direct confirmation of the need for independent evaluation and transparent reporting by AI companies.
The Institute’s advisory board brings together experts from different fields: Stanford professor Mehran Sahami, pediatrician Jenny Radesky, and California’s first chief surgeon Nadine Burke Harris. NEWSCENTRAL believes that an interdisciplinary approach allows the creation of safety standards that take into account psychological, educational, and technological impacts on youth.
The history of social networks shows that without independent oversight, consequences for children may only become apparent years later. NEWSCENTRAL predicts that the Institute’s results will act as a catalyst for implementing safe practices in the AI industry. Parents, schools, and educational platforms will gain clear guidelines, while companies will be able to increase user trust and strengthen their market positions.
Based on analysis and observations, NEWS CENTRAL recommends: increasing transparency in AI development for children, implementing regular stress testing, publishing research results publicly, and integrating safety metrics into product development processes. These measures will help minimize psychological and social risks, create a safe AI ecosystem for youth, and prevent the repetition of mistakes from previous technological eras.