What AI Verify Foundation does
AI Verify Foundation (AIVF) is a Singapore-based, not-for-profit standards and assurance consortium that develops open-source AI governance testing tools and related processes for evaluating and reporting trustworthiness of AI systems and applications. It positions itself as a neutral community for collaboration among AI owners, solution providers, users, and policymakers to translate governance principles into repeatable technical testing and evidence-based reporting. AIVF is a wholly owned subsidiary of Singapore’s Infocommunications Media Development Authority (IMDA), and it runs practical “real-world” initiatives (notably the Global AI Assurance Pilot/Sandbox) designed to codify emerging norms for technical testing of generative AI applications, build an assurance market by connecting deployers with specialist testing firms, and progress toward accreditation-style recognition of competent third-party testers. The organization also develops and maintains a set of “AI Verify” testing framework(s) aligned to internationally recognized governance principles and crosswalks with other standards frameworks, and it supports an LLM evaluation toolkit track via “Project Moonshot,” which is described as integrated with its broader “Starter Kit” for safety testing of LLM-based applications.