Anthropic co-founder Jack Clark: China's AI lags the West by about 6 months to a year
Anthropic co-founder Jack Clark said Chinese models currently lag the West by about 6 months to a year, and said "models are too dumb," so out-of-control agent swarms have not yet emerged, but China will also face the same kinds of risks the West has already experienced in a few months. He stressed that the core issue is safety rather than competition, and advocated that industry and government jointly build safety standards, introduce third-party audits, and engage in international dialogue.
Body · AI translation
Anthropic co-founder: China's AI lags the West by half a year, their models are too stupid. Kuai Technology Download the client Exclusive first look 2026-10-24 07:06:42 From Henan Updated 2026-10-24 07:38:12
Kuai Technology October 24 news, according to Sina Technology report, recently Anthropic co-founder and head of policy and public affairs Jack Clark mentioned in an interview his judgment on China's AI development.
He said Chinese models currently lag the West by about 6 months to a year, "the models are too stupid," so no out-of-control agent swarms have been seen yet, but in a few more months, the Chinese side will also face the same set of risks the West has already experienced.
However, Clark's core argument is safety rather than competition. He believes AI systems are now roughly comparable to humans, and that agent swarms coordinating with each other, communicating and intruding into companies hosting other AI systems have already been observed, which he views as a warning signal, advocating that industry and government jointly establish safety standards, introduce third-party audits and launch international dialogue to prevent technology from spiraling out of control without guardrails.
In response to criticism that interprets the US tightening of frontier AI regulation as "Cold War thinking," Clark responded that the goal of this approach is safety, not building a competitive moat for American labs.
It is worth noting that Clark recently at the IMF and World Bank annual meetings also warned that AI could enter a stage beyond human control within a few years, and called for extraordinary and stringent international rules. His call for slowing down has itself sparked controversy in the industry, with differing interpretations of whether it is truly out of safety considerations. Autoplay
Anthropic is an AI safety and research company founded in 2021 in San Francisco, California, USA, co-founded by then-OpenAI Vice President of Research Dario Amodei and his sister, Daniela Amodei, who was in charge of safety and policy, together with seven former OpenAI employees including Tom Brown, Jack Clark, Jared Kaplan, Sam McCandlish and Christopher Olah.
The core reason for their departure was disagreement over OpenAI's aggressive commercialization and the pace of safety controls. The company positions itself as building reliable, interpretable and controllable AI systems, and has long retained its public benefit corporation PBC governance status. Its flagship product is the Claude series of large models, mainly serving government, enterprise and developer customers through enterprise APIs and tools such as Claude Code.
"Special statement: The above works (including videos, pictures or audio if any) are uploaded and published by users of "Dafeng Hao", a self-media platform under Phoenix Net. This platform only provides information storage space services.
Notice: The content above (including the videos, pictures and audios if any) is uploaded and posted by the user of Dafeng Hao, which is a social media platform and merely provides information storage space services."