Sources: Google, OpenAI, and Anthropic plan to launch an AI safety standards body without government oversight in late 2026 or 2027, and are considering CEOs (The Information) | TickerVault