読書リサーチレーダー投資フレームワーク
ログイン / 新規登録
← 読解に戻る
ニュース重大1 分で読めるDisconfirmAI 編集部

Google, OpenAI and Anthropic reportedly move toward a joint frontier-AI standards body

The Information reports the three labs are shaping SAFA and discussing cross-testing of each other's commercial models; no official confirmation yet.

2026-09-24 19:43 GMT+8

According to The Information, Google, OpenAI and Anthropic are closer to setting up an independent body to develop safety standards for frontier AI, tentatively called the Standards Authority for Frontier AI (SAFA).

The report says the body would operate independently and address gaps in government oversight. The companies have also discussed an arrangement to test each other's commercially available AI models for vulnerabilities and unexpected behaviour, while restricting retention of data obtained during those evaluations.

No company has officially confirmed the plan, and the name is tentative. The same day, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei addressed the UN Security Council on AI safety: Altman said major AI decisions should be shaped through democratic institutions and governments, while Amodei called for international agreements on testing new models.

ソース

  • proactiveinvestors.com↗
リサーチを続ける →
ログイン / 新規登録
読書リサーチレーダー投資フレームワーク