The Pause Giant AI Experiments letter was an open letter published on March 22, 2023 by the Future of Life Institute, calling on AI labs to pause training of systems more powerful than GPT-4 for at least six months. It gathered more than 30,000 signatures, including prominent technologists and researchers, and became one of the most widely discussed public interventions in AI policy debates of the 2020s, though it did not lead to any actual pause by major labs.
Content and signatories
The letter argued that AI labs were locked in an "out-of-control race" to develop and deploy increasingly powerful systems that no one, including their creators, could reliably understand, predict, or control, and warned of risks ranging from the flooding of information channels with propaganda and untruth to broader loss of control over civilization. It called for a public, verifiable pause, and for the intervening time to be used to develop shared safety protocols overseen by independent experts. Signatories included Max Tegmark, whose Future of Life Institute organized the letter, Elon Musk, and numerous academics and figures from the AI safety community, though notably few researchers from the very top frontier labs signed.
Reception and criticism
The letter drew sharp criticism from multiple directions. Some AI safety researchers, including figures associated with existential risk concerns, argued a temporary pause was an insufficient response to the risks described and preferred more binding regulatory measures. Other critics, including scholars such as Timnit Gebru and Emily Bender, argued the letter focused on speculative long-term harms while distracting from documented near-term harms such as algorithmic bias, and questioned the motives of signatories with commercial interests in slowing competitors.
Aftermath
No major AI lab paused development in response to the letter. OpenAI continued shipping products, and competitors including Google DeepMind and Anthropic continued releasing new models in the following months. The letter is nonetheless credited with elevating public and political attention to AI risk, contributing to momentum that fed into subsequent policy efforts including the Bletchley Park AI Safety Summit later that year and broader discussion of AI governance frameworks. It is frequently referenced as a marker of how mainstream concern about frontier AI capability had become by early 2023, even as its central request went unmet.