Garry Tan urges U.S. AI labs to distill frontier models as public good

Frontier artificial‑intelligence models, which are built on vast datasets drawn from publicly available human knowledge, are being championed by researcher Tan as a resource that ought to be treated as a public good. In a recent commentary, Tan emphasized that because the training material for these models originates from open sources, the resulting capabilities should be broadly accessible rather than confined to a limited set of commercial entities.

Tan’s argument underscores a growing debate over the governance of advanced AI systems, highlighting the tension between proprietary control and societal benefit. By framing access to capable AI as a public good, the stance calls for policies that promote open availability, collaborative oversight, and equitable distribution, aiming to ensure that the transformative potential of frontier models serves the wider public interest.