U.S. tech groups urge Washington to avoid curbs on open AI models

Traders Union
All News
Business & Wealth
US tech groups on open AI curbs

Highlights

Nvidia, Palantir, Microsoft and Meta urged Washington in a joint letter Friday to avoid premature curbs on open-weight AI models amid China-related security concerns.
OpenAI and Anthropic, each nearing $1tn valuations in 2024, maintain closed AI model strategies while investors accuse them of overstating open-source risks to secure market dominance.
A preliminary U.S.-UK study finds China’s Kimi K3 model behind top American models, with Nvidia CEO calling for open-source embrace and renewed chip sales to China to uphold U.S. AI leadership.

Industry pushback against possible restrictions

As reported by Financial Times, Nvidia, Palantir, Microsoft and Meta are among the groups urging the U.S. not to restrict open-weight AI models as officials weigh tougher action against Chinese labs. Their joint letter says policymakers should keep the AI frontier plural by avoiding premature limits on open models, which users can adapt and run on their own hardware.

The signatories do not include OpenAI, Anthropic or Google, the main developers of closed leading models. The debate intensifies after the release of Kimi K3 by Beijing-based Moonshot AI, which raises concerns in Washington that China could narrow the gap with American rivals in a field both countries treat as strategically important.

Michael Kratsios, director of the White House Office of Science and Technology Policy, says the U.S. has information that Moonshot used distillation from Anthropic's latest Fable model. Anthropic says industrial espionage creates serious national security risks for the United States and democratic allies, while Treasury secretary Scott Bessent says sanctions are on the table when Chinese firms cross the line into intellectual property theft, although he adds that the Trump administration supports open-weight AI.

Competitive and market stakes for AI developers

The dispute also reflects a business model clash inside the U.S. AI sector. OpenAI and Anthropic reach valuations close to $1tn this year as investors bet they can preserve a lead over foreign rivals and continue charging premium prices for their models.

The Friday letter argues that open models help businesses avoid paying frontier-model prices for every task and can widen AI adoption. Investors including former Trump AI adviser David Sacks accuse Anthropic of overstating the danger from open-source competitors in order to protect its own market position.

A preliminary assessment released on Thursday by U.S. and UK researchers finds Kimi K3 still trails the top American models, while U.S. open-weight models lag further behind. In an Axios interview this week, Nvidia chief executive Jensen Huang says open-source models, including Chinese ones, should be embraced by U.S. companies, and he continues to press Washington to let Nvidia sell AI chips in China, arguing that wider access would reinforce American leadership.

The companies behind Friday's letter, including Andreessen Horowitz, IBM and Reflection, acknowledge that open-weight systems carry distinct risks because outside users can modify them after release. Still, they argue that blocking such models would increase those risks rather than reduce them, because broader access allows researchers and developers to test behavior, identify vulnerabilities and build safeguards over time.

Our earlier report on the industry letter opposing restrictions on open-weight AI models explained how Nvidia, Microsoft, Meta, Palantir and other firms urged U.S. policymakers to avoid premature limits that could curb competition and slow innovation. The piece also outlined how Moonshot AI’s Kimi K3 intensified Washington’s focus on China’s progress and allegations of unlawful model distillation, with executives and officials arguing for targeted legal responses rather than sweeping bans.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.