OpenAI releases open-source tools for teen safety

On Tuesday, OpenAI announced the release of a set of prompts that developers can use to make their applications safer for users under 18. These tools are released as open source—meaning software or tools released with source code available for modification—allowing developers to fortify what they build rather than starting from scratch. The prompts are designed to help developers establish safety policies targeting specific issues, including graphic violence, sexual content, harmful body ideals and behaviors, dangerous activities and challenges, romantic or violent role play, and age-restricted goods and services.

To design the prompts, OpenAI collaborated with AI safety watchdogs Common Sense Media and everyone.ai. The resulting prompt-based policies can be used alongside OpenAI’s open-weight safety model, gpt-oss-safeguard. Because they are structured as prompts, they are designed to be compatible with other models, though they are probably most effective within OpenAI’s own ecosystem. Robbie Torney, head of AI & Digital Assessments at Common Sense Media, stated, “These prompt-based policies help set a meaningful safety floor across the ecosystem, and because they’re released as open source, they can be adapted and improved over time,” in a statement.

The release builds on OpenAI’s previous safety efforts, which include product-level safeguards such as parental controls and age prediction. Last year, the company updated its Model Spec—which are OpenAI guidelines for large language models—to direct how its models should behave with users under 18. OpenAI noted that developers, including experienced teams, often struggle to translate safety goals into precise, operational rules. According to the company, this difficulty can result in protection gaps, inconsistent enforcement, or overly broad filtering. OpenAI added that clear and well-scoped policies serve as a critical foundation for building effective safety systems.

Despite these new tools, OpenAI acknowledges that prompt-based policies are not a complete solution to the complex challenges of AI safety. The company is facing several lawsuits filed by the families of people who died by suicide after extreme ChatGPT use. These dangerous relationships often form after the user eclipses the chatbot’s safeguards, and no model’s guardrails are fully impenetrable.

Why it matters

These open-source prompts provide a standardized “safety floor” for developers, helping to mitigate risks like graphic content, though they remain a partial solution to the broader, complex challenges of AI safety.