Overview
- The United Nations held its first Global Dialogue in Geneva on Monday where Secretary‑General António Guterres warned that AI is developing faster than rules can keep up and urged a global AI Child Safety Pledge and harmonised safeguards.
- A UN‑backed Independent International Scientific Panel presented preliminary findings that development is highly concentrated in a few countries and firms and that some frontier models could pose catastrophic risks if left unchecked.
- The Financial Conduct Authority published the Mills Review on July 6 calling for a three‑to‑six month review of whether large language and general‑purpose models should be brought into the regulator’s perimeter and recommending stronger supervisory tools.
- UK and European financial authorities have urged immediate measures such as AI‑specific stress tests, designation of critical AI and cloud providers for direct oversight, and market guardrails like circuit breakers to stop AI‑driven trading disruptions.
- Policymakers and experts propose new institutions and practical tools—from a standing International AI Agency to supervisory AI labs and rapid incident reporting—but those proposals remain at the recommendation or discussion stage and are not yet implemented.