OpenAI pledges $2 million for research on AI and children’s safety
OpenAI pledged $2 million to a coalition-backed research network developing shared tools to measure AI risks and opportunities for children.
OpenAI pledged $2 million to support a scientific research network focused on artificial intelligence and children’s safety, the company announced on November 12. The network is being established under the Coalition for Children’s Rights and Protection in the Age of AI.
The commitment is intended to support shared measures and open-source tools, including classifiers and graders, for technical teams and independent researchers to test and validate. The coalition’s Charter of Engagement for Safer AI for Children says the tools are meant to make research comparable across products, countries and time. OpenAI said the work could help researchers assess risks and opportunities for young people and inform safer products and better policy. Those benefits have not yet been demonstrated.
The company said it signed the charter earlier that week at the Paris Peace Forum. The document says the coalition is led by France, Spain and Kenya, involves 10 United Nations agencies, and brings together governments, researchers, civil society and industry.
Under the planned process, companies may run shared measurement tools within their own systems and provide coded results, reducing the need to transfer full conversations. The charter assigns scientific priorities, methods, analysis and conclusions to independent scientific responsibility. It also requires publication of methods and results, including null, negative or inconclusive findings, and says financial or in-kind contributors will not receive a veto over publication.
The charter limits industry funding to the Shared Measurement workstream. It says other parts of the programme—including the Observatory, research on evidence and developmental outcomes, Scientific Committee remuneration, and assessment and standards—must be funded exclusively through public and independent philanthropic sources, with separate accounting. The opened materials do not identify who will administer OpenAI’s payment, when funds will be transferred or when the tools will be delivered.
More news

Florida seeks court limits on OpenAI model development and ChatGPT behavior

OpenAI endorses California youth AI safety bill SB 1119

More than 1,200 OpenAI and Anthropic staff sign 'Pacing the Frontier' statement
