OpenAI Strengthens Teen Safeguards in ChatGPT with New Under-18 Principles
New under-18 principles added to Model Spec
Age-based safeguards and parental controls
New AI literacy resources for families
OpenAI Introduces New Safeguards for Teen Users of ChatGPT
In a significant move to enhance the safety and well-being of younger users, OpenAI has introduced new safeguards specifically tailored for teenagers using ChatGPT. This update to the Model Spec introduces under-18 principles aimed at ensuring safety, transparency, and parental oversight for users aged 13 to 17. As schools, families, and policymakers increasingly scrutinize the role of generative AI tools in education and daily life, these changes signal a commitment to age-appropriate AI design.
A Response to Growing Concerns
In a recent LinkedIn post, Chris Lehane, OpenAI’s chief global affairs officer, emphasized the shared concerns among parents, educators, and policymakers about providing "safe, trustworthy AI at home and at school." The Model Spec acts as the authoritative guideline for how the AI is expected to operate, evolving in response to societal needs and expectations.
The updated Model Spec introduces a dedicated set of principles organized around four key commitments:
- Putting Teen Safety First: Prioritizing the well-being of users aged 13 to 17.
- Promoting Real-World Support: Offering guidance and alternative resources in high-risk situations.
- Treating Teens as Teens: Acknowledging the unique needs and contexts of teenage users.
- Setting Clear Expectations Through Transparency: Ensuring users understand AI interactions and limitations.
These principles are particularly vital as today’s teenagers represent the first generation to grow up with AI as a constant companion in education, communication, and preparation for the workforce.
Enhanced Safeguards and Parental Controls
The recent updates also include expanded practical applications of teen safeguards. OpenAI is rolling out age prediction technologies to help automatically apply under-18 protections to accounts suspected to belong to minors. When a user’s age is uncertain, the default settings will direct them toward an under-18 experience, allowing parents to verify their age for full access.
Moreover, these safety measures have been incorporated across new products such as group chats, the ChatGPT Atlas browser, and the Sora app. Existing parental controls empower families to manage settings, including quiet hours, content controls, and notifications for acute distress situations.
Lehane noted that OpenAI is also keen on engaging with policymakers to further advance teen safety, using the updated Model Spec as a foundation for broader discussions on the responsible use of AI.
New AI Literacy Resources for Families
Accompanying the Model Spec update are two newly launched AI literacy resources designed for teens and parents. The first is a family-friendly guide that explains how AI models are trained, their propensity for mistakes, and provides practical advice on using tools like ChatGPT more responsibly. This guide includes tips on crafting effective prompts, verifying information, managing data, and comprehending AI limitations.
The second resource serves as a toolkit for parents, featuring conversation starters and advice on fostering critical thinking, setting boundaries, and discussing sensitive or emotional topics. Developed with input from OpenAI’s Expert Council on Well-Being and AI and partners like ConnectSafely, these resources aim to equip families with the knowledge they need for engaging with AI.
Conclusion
With these new safeguards and resources, OpenAI is demonstrating its commitment to ensuring that safety evolves alongside youth engagement with technology. As young minds navigate their formative years in a world increasingly influenced by AI, it is crucial to prioritize their safety and education. "This is the first generation growing up with AI,” Lehane states, “Our job is to make sure safety keeps pace with how these tools are actually used.” In doing so, OpenAI is not just responding to current concerns but proactively shaping a safer and more aware environment for future generations.