🚨 NEW: Anthropic, OpenAI and Google have quietly held working-group talks since July on building their own AI safety standards body, per The Information. (Cointelegraph)