The Statement on AI Risk is a one-sentence open letter published by the Center for AI Safety (CAIS) (CAIS) on May 30, 2023, at https://aistatement.com/. It equates the risk of human extinction from AI with other societal-scale risks such as pandemics and nuclear war, and was signed by more than 350 AI researchers, executives, and public figures.
The statement
The entire substantive text is a single sentence:
"Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."
CAIS designed the statement to be minimal in order to maximize signatory breadth. It carries no specific policy demand, advancing only a framing claim about the magnitude of the risk.
Key claims
Two claims are central to how the statement has been read. The first is that extinction-level AI risk is a mainstream concern rather than a fringe one, a reading demonstrated by the composition of the signatories (confidence: high). The second is that catastrophic AI risk should be treated as a global priority on par with pandemics and nuclear war. This second framing is itself disputed: critics argue that near-term harms deserve priority instead, a reception discussed at FLI — Pause Giant AI Experiments: An Open Letter (confidence: contested).
Signatories
Signatories included AI researchers Geoffrey Hinton, Yoshua Bengio, Stuart Russell, Ian Goodfellow, and Ilya Sutskever; frontier-lab chief executives Sam Altman (OpenAI), Dario Amodei (Anthropic), and Demis Hassabis (Google DeepMind); industry figures Kevin Scott (Microsoft CTO) and Bill Gates; and public figures including Audrey Tang and Laurence Tribe. Unlike the FLI "Pause" letter two months earlier, all three major US frontier-lab CEOs signed.
Reception and use
The statement has been widely cited in subsequent US and international AI governance to establish that existential and catastrophic AI risk is a mainstream concern among technical experts. It has been used in advocacy around EO 14110, the Bletchley Declaration (The Bletchley Declaration (AI Safety Summit, 1–2 November 2023)), and California SB 1047 (California SB 1047 — Safe and Secure Innovation for Frontier AI Models Act (enrolled + veto)). Within the AI-safety-as-existential-risk current that shaped the Biden-era regulatory push and the FLI pause tradition, the statement's brevity stands in contrast to FLI's more detailed and prescriptive pause letter; CAIS's brevity was a strategic choice to broaden the signatory base.
Relationships
- related: FLI — Pause Giant AI Experiments: An Open Letter — precursor (March 2023) with a more prescriptive framing
- supports: AI Autonomy Risk, Compressed 21st Century — provides high-profile expert endorsement of catastrophic-risk framing
- depends-on: Center for AI Safety (CAIS) — publishing organization
- related: Executive Order 14110 — Safe, Secure, and Trustworthy AI, The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), California SB 1047 — Safe and Secure Innovation for Frontier AI Models Act (enrolled + veto) — downstream governance documents citing this statement
- related: PauseAI — activist descendant of the same advocacy current