Longitudinal evidence needs to precede confident policy
OpenAI announced on September 8 that it will commit $5 million to independent research on how generative AI affects people ages 13–17. Priorities include emotion regulation, identity and self-esteem, friendships and family relationships, differences by culture, language and socioeconomic context, and the effectiveness of age-appropriate design, AI literacy and safeguards.
The call is international and accepts qualitative, quantitative, experimental, observational and participatory methods. Studies involving minors must address institutional or independent ethics review, parental consent and age-appropriate assent, privacy and sensitive data, disclosures of harm and relevant child-development expertise.
Funding research does not guarantee an independent conclusion
This is a grant announcement, not evidence that AI is beneficial or harmful to teens. Because the funder also makes the product, actual agreements should be examined for preregistration, outcome definitions, publication of negative findings, data access and editorial independence. OpenAI requires applicants to disclose independence and conflicts, but transparency after selection and publication will matter as much as the application form.
Schools and regulators should continue applying existing child-safety and privacy duties while the evidence develops, and avoid treating one program's future findings as universal across ages, countries and patterns of use.