Nvidia vs. Anthropic: The open-weight AI schism engineering leaders can't ignore

Nvidia CEO Jensen Huang made his first X post to publish an industry letter supporting open-weight AI models, signed by nearly every major AI company except Anthropic. The letter argues that open-weight models strengthen competition, enable sovereign AI, and are critical for security through transparency. It directly responds to US government discussions about potentially banning

Chinese open-weight models, framing such restrictions as harmful to American AI leadership. Anthropic CEO Dario Amodei responded with his own letter, clarifying that Anthropic does not advocate for banning open-weight models. Instead, he argues for three policy priorities: keeping advanced chips out of authoritarian hands, cracking down on industrial-scale distillation, and requiring mandatory safety testing

for all sufficiently capable models regardless of origin. He directly disputes the letter's claim that open weights necessarily help defenders more than attackers. The core tension is about distillation—the practice of using one model's output to train another. Anthropic views large-scale distillation from Chinese companies as a national security threat that enables authoritarian states to

catch up. The industry letter reframes distillation as a legitimate innovation technique, akin to open-source software traditions, that should not be conflated with theft. Engineering leaders face a strategic choice: the open-weight ecosystem enables cost flexibility and avoids vendor lock-in but introduces governance challenges around model safety and data sovereignty. Anthropic's position appears self-interested but