Anthropic and Nvidia Unite Against Open-Weight Bans, Shifting the US AI Policy Debate
A safety-focused lab and the dominant AI chip supplier aligning on open-weight policy hands regulators a new frame: target infrastructure, not model weights.
3. Anthropic and Nvidia Unite Against Open-Weight Bans, Shifting the US AI Policy Debate
On July 28, 2026, Anthropic CEO Dario Amodei published a detailed position statement arguing that open-weight models without dangerous capabilities are "a public good," while Nvidia CEO Jensen Huang shared an open letter on X co-signed by IBM, Microsoft, Meta, Mozilla, and others, calling open-weight models essential to innovation, competition, and national sovereignty. Both parties rejected blanket prohibitions, redirecting attention toward export controls on AI training and inference silicon, restrictions on industrial-scale distillation, and mandatory safety testing for sufficiently powerful models regardless of whether they are open or closed. The statements follow U.S. Treasury Secretary Scott Bessent signaling that the White House could sanction open-weight Chinese frontier labs, and OSTP Director Michael Kratsios accusing Moonshot AI of using distillation to train Kimi K3.
The alignment between Anthropic and Nvidia is strategically significant precisely because the two companies have different incentive structures. Anthropic competes directly with open-weight providers like Meta's Llama and Alibaba's Qwen; Nvidia sells the silicon that runs all of them. Their convergence on the same policy frame gives regulators political cover to move away from model-level bans and toward the infrastructure controls Anthropic specifically named: chip export restrictions and anti-distillation rules. That framing also, notably, protects Anthropic's own frontier models Mythos and Fable from a regulatory regime that could force disclosure of weights they have no intention of releasing, while still constraining Chinese labs through hardware supply chain pressure.
The distillation concern is the thread worth pulling. Kratsios named Moonshot AI's Kimi K3 as a suspected distillation product, and Anthropic flagged that its own models may have been used to train foreign systems without authorization. If the U.S. moves toward enforceable anti-distillation rules, the next question is verification: how regulators would detect industrial-scale distillation without access to training logs. That enforcement gap may define the next round of policy debate.
Source: Anthropic and Nvidia come out against blanket bans on open-weight AI models