OpenAI commits up to $5 million to independent research on AI and teen development

OpenAI committed up to $5 million to a grant program for independent research into how generative AI affects the lives and development of people ages 13 to 17.

Applications opened September 8 and close October 6, 2026. OpenAI said individual awards may reach $1 million and selected applicants will be notified on or before November 13. OpenAI Group PBC, not the OpenAI Foundation, will fund and administer the program.

The program is designed to build evidence, not to establish that AI has any particular effect on teenagers. OpenAI is seeking studies on patterns of generative-AI use, social and emotional development, differences across ages and contexts, and whether safeguards, product design, AI literacy or support from adults affect outcomes. Eligible methods include qualitative, quantitative, experimental, observational, theoretical, participatory and mixed-methods research.

Proposals involving minors must describe their institutional or independent ethics-review process, informed consent and age-appropriate assent, privacy and data protections, and procedures for responding to disclosures of harm or imminent risk. Applicants must also address sensitive personal, behavioral or mental-health information, relevant child-development or safeguarding expertise, and parental consent where appropriate. An ethics and safeguarding statement is required for research involving minors, sensitive information or potentially vulnerable populations. Those participant protections address a different context from OpenAI’s privacy-preserving misuse-detection system for API customer prompts.

Applicants must disclose relevant financial relationships, industry funding and other potential conflicts. OpenAI lists “independence and credibility” as a selection criterion, including whether a project can produce trustworthy findings regardless of whether they favor AI providers. Those requirements describe how proposals will be reviewed; they do not establish that funded work will be free from funder influence. Neither the announcement nor the application portal publishes a grant agreement detailing OpenAI’s rights over study design, data access or publication review, and neither names the internal experts or advisers who will select recipients.

Each funded project must provide OpenAI with a final report covering its methods, evidence, limitations, uncertainty and any supported recommendations, plus an interim progress update in the first quarter of 2027. OpenAI encourages peer-reviewed papers, preprints or public reports, but public release is not a condition of funding.

As an external benchmark, U.S. health-research rules require institutional review board oversight and provisions for children’s assent and parental or guardian permission for covered research involving children. Those rules apply to research conducted or supported by the U.S. Department of Health and Human Services, so they do not automatically govern every project in OpenAI’s program. UNICEF’s ethics policy separately treats ethics review, safeguarding, risk management, consent or assent, privacy and data management as lifecycle requirements for UNICEF-supported evidence work; it is a benchmark, not a rule for OpenAI’s grants.

Applicants must be at least 18 and either affiliated with a research institution or organization or have significant relevant experience. OpenAI said it will not prioritize for-profit organizations.