Q: What is the Youth AI Safety Institute and what will it do? A: The Youth AI Safety Institute is an independent research and testing lab launched by Common Sense Media to study risks AI tools may pose to children and teens. It will red-team popular models used by young people, publish consumer-friendly ratings and safety benchmarks, and advance AI safety testing for kids to help families choose safer tools. Q: Who is funding the Youth AI Safety Institute and will funders influence its research? A: The institute starts with a $20 million annual budget backed by the OpenAI Foundation, Anthropic, Pinterest, the Walton Family Foundation, Goldman Sachs Managing Director Gene Sykes, and other philanthropists. According to Common Sense Media, funders will have no say in the group’s operation or research. Q: What specific harms and risks will the lab test for in AI tools? A: The lab will test for harms such as self-harm or suicide advice, sexualized or suggestive content including digital undressing, violent or hateful outputs, and problematic data practices like unclear collection or sharing. AI safety testing for kids will also assess design flaws such as missing age verification, weak moderation, and the absence of in-product crisis resources. Q: How will the institute publish its findings and safety benchmarks? A: The institute plans to publish consumer-friendly reports, simple ratings, and youth safety benchmarks that tech companies can use, and it intends to start releasing research this month. It also plans to update findings often so benchmarks keep pace with frequent model updates. Q: What are the red flags parents can use to spot risky AI apps right now? A: Red flags include no way to set or verify age, suggestive or violent outputs slipping through, chatbots that refuse too little or too late on sensitive prompts, and hidden or vague moderation policies. Parents should also watch for missing crisis resources, image-editing deepfakes, and unclear data or privacy practices. Q: How should schools and youth programs use upcoming ratings and benchmarks? A: Schools and youth programs should require evidence that a tool has passed independent youth safety benchmarks and ask vendors how they block self-harm, sexual, and violent content and whether they disable training on student data. They should use AI safety testing for kids ratings to choose tools that meet benchmarks, re-check scores after major updates, and demand audit access and timely safety fixes. Q: How will the institute keep up with the rapid pace of AI model updates? A: Because AI models often gain new update capabilities weekly or monthly, the institute is set up as a separate team to run frequent, rigorous research and red-team leading models as they change. It plans to update findings often and publish revised benchmarks so safety assessments can keep pace with rapid development. Q: How can families apply ratings and benchmarks when choosing apps for their children? A: Families should use AI safety testing for kids results like crash-test stars to filter choices before a child downloads a tool and favor apps that meet or exceed youth safety benchmarks. They should pair ratings with quick at-home tests, re-check scores after big updates, and ask vendors for proof of third-party testing and details about past fixes.