Google, OpenAI, Anthropic plan AI safety standards group by 2027

Google: Google, operating through its parent company Alphabet, develops and deploys advanced artificial intelligence technologies via divisions including Google DeepMind. It is actively collaborating with OpenAI and Anthropic to launch an independent standards body for frontier AI models. OpenAI: OpenAI is an AI research organization focused on building advanced systems with emphasis on safety, alignment, and responsible deployment. It is partnering with Google and Anthropic to establish the Standards Authority for Frontier AI as a private-sector initiative for model safety standards and testing. Anthropic: Anthropic is an AI company dedicated to creating safe, reliable, and interpretable artificial intelligence systems. It is working alongside Google and OpenAI to advance the proposed independent AI safety standards group targeted for launch in late 2026 or early 2027. Standards Authority for Frontier AI: The Standards Authority for Frontier AI is a planned independent organization that will set safety and security standards for frontier AI models, support third-party testing, and establish rules for incident reporting and auditor qualifications. It is being developed by Google, OpenAI, and Anthropic outside of government oversight following stalled public-private efforts. Frontier AI Focus: The initiative targets safety and capability evaluation for the most advanced AI models through independent auditing and incident reporting mechanisms. Industry Self-Regulation: Major AI developers are coordinating on voluntary safety standards and testing protocols in the absence of a public-private regulatory framework.