OpenAI backs new foundation for AI standards
Appia Foundation helps define open AI assessment standards for safer governance See how shared testing criteria could boost trust across the AI supply chain
OpenAI said it helped found the Appia Foundation, a Linux Foundationhosted group that will develop open, modular specifications for evaluating and governing advanced AI systems. The organization aims to help align international standards with practical testing criteria across the AI supply chain.
The company said the effort is meant to improve trust in AI assessments by giving governments, companies, and thirdparty evaluators a shared technical language. OpenAI linked the initiative to its broader safety and governance work, including its frontier safety blueprint, thirdparty evaluation guidance, and testing partnerships with US and UK AI safety institutes.
According to OpenAI, stronger national institutions and international cooperation will both be needed to manage frontier AI risks. Appia’s work is intended to make assessment practices more interoperable across organizations and jurisdictions.