Global AI Governance
From Bletchley to New Delhi
The global AI governance architecture has evolved through a series of high-level summits since 2023. The first AI Safety Summit was held at Bletchley Park, United Kingdom, in November 2023, producing the Bletchley Declaration signed by 28 countries (including the US, China, EU, and India), which identified AI safety risks — especially from frontier AI models — as a shared global responsibility. The second summit was held in Seoul, South Korea (May 2024), the third in Paris, France (February 2025), and the fourth — the India AI Impact Summit — in New Delhi (February 2026). Each summit has broadened participation and shifted emphasis from pure safety concerns toward balancing innovation with risk management, and latterly toward AI's developmental potential for the Global South.
- AI Safety Summit 1: Bletchley Park, UK, November 2023 — 28 signatories to Bletchley Declaration; focused on frontier model risks.
- AI Safety Summit 2: Seoul, South Korea, May 2024 — adopted Seoul Declaration; focus on AI safety research and international cooperation.
- AI Safety Summit 3: Paris, France, February 2025 — broadened to AI for public benefit.
- AI Safety Summit 4 / India AI Impact Summit: New Delhi, February 16–21, 2026 — 88 countries; New Delhi Declaration; "AI for All" framing.
- Global Partnership on AI (GPAI): Multilateral body (founded 2020) to guide responsible AI development; India is a founding member.
- EU AI Act (2024): First comprehensive binding AI regulation globally — risk-based framework; prohibits unacceptable-risk AI (social scoring, real-time biometric surveillance); regulates high-risk AI (healthcare, critical infrastructure, biometrics); India has not enacted equivalent binding law.
● Tracked since February 16, 2026 · last seen June 15, 2026 · updates as the daily brief publishes