OpenAI Pushes Global Youth AI Safety Standards at G7

OpenAI is urging global leaders to establish international standards for youth AI safety, proposing the creation of a dedicated institute to oversee these efforts. The company plans to discuss these initiatives at the upcoming G7 Summit in Évian, France.

AI offers significant potential for young people, from personalized learning to career preparation. However, realizing these benefits requires ensuring AI is accessed safely and is age-appropriate. OpenAI argues this responsibility extends beyond parents and should be embedded by companies through robust safeguards.

The proposed international institute would foster continuous collaboration among governments, researchers, civil society, and industry to share evidence and develop practical guidance. This mirrors existing efforts, such as Common Sense Media’s Youth AI Safety Institute, which OpenAI supports.

Setting a Global Standard

OpenAI outlined key principles for a strong youth AI safety framework. Companies should identify minors and implement age-appropriate protections, defaulting to safeguards when a user's age is uncertain. This requires effective, privacy-preserving age estimation methods.

Regular risk and benefit assessments for young users are crucial. These evaluations must consider developmental stages and empirical evidence to proactively address potential harms and promote positive outcomes like learning and creativity.

Accessible parental controls are essential, allowing guardians to manage settings like memory, data usage, and time limits. Companies must actively promote these tools and provide transparent information about AI's benefits and risks for young people.

Clear protocols are needed for serious safety situations, including exploitation and harmful content. AI systems should be designed to prevent the generation of unsafe or developmentally inappropriate material for children.

Furthermore, AI should supplement, not replace, human judgment and real-world relationships. The personal information of minors must be protected, prohibiting privacy-invasive advertising and the sale of their data.

The company emphasizes that youth AI safety frameworks should promote opportunity and literacy alongside protection. This includes equipping young people with the knowledge to use AI safely and responsibly, fostering critical thinking skills.

Strong accountability mechanisms, such as independent audits based on common standards, are vital. Legislative frameworks should provide oversight and enforcement to ensure companies comply with safety and privacy obligations.

OpenAI states these principles are reflected in ChatGPT's design, including strengthened safeguards for users under 18, parental controls, and advanced age-prediction systems. When a user's age is uncertain, the system defaults to stronger protections.

The company is actively engaged in research-driven deployments and teacher training through initiatives like Education for Countries, partnering with nations such as Estonia, Greece, and Singapore.

Beyond the G7 Leaders’ Summit, OpenAI will host the OpenAI Forum in Paris to facilitate practical discussions on translating shared youth safety goals into concrete safeguards and standards.