API
< Back to all clusters
[TECHNOLOGY] · United States · 40 sources

started · updated

Demis Hassabis Calls for US‑Led Global AI Standards Body

DeepMind chief Demis Hassabis, a Nobel laureate, has urged the United States to create a new public‑private AI standards authority modeled on the Financial Industry Regulatory Authority (FINRA). The body would develop assessment protocols and benchmarks for frontier AI models, testing them for cyber‑security, biological and national‑security risks before public release. Initial testing would be voluntary, with a 30‑day pre‑release window, potentially becoming mandatory for market entry. Funding would come largely from the industry to secure world‑class expertise and compute resources. Hassabis argues the rapid progress toward artificial general intelligence could have impacts ten times larger and faster than the industrial revolution, making a proactive oversight framework essential. He notes the US’s economic and technical position makes it well‑placed to lead, and references recent calls from Nobel laureates and economists warning of AI‑driven economic disruption. The proposal has drawn mixed reactions, with some analysts questioning the effectiveness of self‑regulation and its international acceptability, but Hassabis maintains it is preferable to post‑deployment regulation.

Sources