AI News Today (Nov 12, 2025): UK approves AI safety testing law, Google unveils ‘Private AI Compute,’ OpenAI opens Delhi office, Foxconn teases OpenAI tie‑up, and $750M floods legal‑AI
The UK has passed a law allowing authorized researchers to legally test AI models for their ability to generate child sexual abuse material, following a doubling of AI-CSAM reports from 199 in 2024 to 426 in 2025. The measure amends crime and policing law to permit pre-emptive safety checks by AI developers and child-protection groups.