On January 13, 2025, the Trump administration formalized a rule permitting Nvidia to export H200 chips to China, with the volume capped at approximately 2 million units annually, equivalent to the total compute currently held by a typical US frontier AI company. The policy reverses the Biden-era export ban on advanced AI chips to China, justified by the argument that permitting strategic access discourages Chinese competitors like Huawei from accelerating independent chip development. Enforcement will rely on know-your-customer requirements to block use by sanctioned entities, though officials acknowledge enforcement challenges. For financial institutions, this materially shifts the addressable-market volatility and revenue-predictability calculus for US chip vendors and hyperscalers, as China-targeted revenue moves from structurally prohibited to conditionally permitted under an administration with signaled willingness to use trade policy as strategic leverage.
360 other tracked developments touching the same claims.