More than 20 tech companies signed an open weight AI letter on July 24 2026. It urges the U.S. government to avoid premature restrictions on accessible artificial intelligence models. The signatories include OpenAI, Nvidia, Meta, Microsoft, and Palantir, alongside infrastructure providers such as Dell, Cisco, Cloudflare, and AMD. The open weight AI letter, published as a policy statement, argues that open-weight models strengthen competition, improve safety, and keep innovation inside the United States. Heavy regulation, the signatories warn, could push development and talent overseas.
Open Weight AI Letter Signers
Nvidia, Microsoft, Meta, and Palantir led the group that organized the open weight AI letter. More than 20 other firms joined them, including Hugging Face, IBM, Mozilla, AMD, and Cloudflare. A Slashdot summary pulled the list together from CNBC reporting. Elon Musk amplified the letter on X with the words “full support,” though his SpaceX company did not formally sign.
OpenAI initially looked absent from the signatory list. Then TechRadar reported that OpenAI quietly added its name after the first publication. Company president Greg Brockman told reporters that “having more models, more usage, that is a good thing.” Sam Altman later posted on X that he wants the U.S. to win with both open-weight and proprietary models and that he was “glad to see this.”
Google and Anthropic are the conspicuous names missing from the open weight AI letter. Both companies keep their most capable models closed. Both have business models built on proprietary access. Their absence lays out the exact split the letter describes: a few firms that control the most advanced closed models, and a broad coalition that depends on a competitive, commoditized AI market.
The group behind the open weight AI letter includes many builders of chips, cloud services, and developer tools. Their businesses do well when thousands of companies can download, customize, and run powerful models without paying per-query fees to a handful of gatekeepers.
What the Letter Asks the Government to Do
The open weight AI letter makes five concrete requests. The first: keep open-weight models legal. The signatories push back against proposals that would require companies to get government approval before releasing AI models that anyone can download and inspect. The letter states that “open-weight models means the AI models that anyone can download, inspect, modify, and run on their own infrastructure are an important part of that foundation because they make advanced AI more accessible, adaptable, and widely available.”
Second, the group asks the government to invest public money in shared AI infrastructure. Think GPU compute for universities and startups, public datasets, and evaluation benchmarks. The goal is simple: make sure smaller players can afford to build, not just rent access from large cloud providers.
Third, the open weight AI letter defends model distillation. Distillation is a common technique where a smaller model learns by studying the outputs of a larger one. The letter warns against broad bans on distillation. It calls instead for targeted legal remedies when someone actually steals intellectual property. Fourth, the letter calls for rules that stop AI from becoming a monopoly controlled by two or three companies. Finally, the companies ask for a seat at the table as policymakers draft new AI regulations.
Satya Nadella, CEO of Microsoft, wrote that “open-weight models are essential to a healthy AI ecosystem” and that the letter outlines a path to “strengthen American competitiveness and expand economic opportunity while protecting national security.” These demands now sit on the desks of lawmakers who are actively writing the next generation of AI rules.
Vested Interests Behind the Unity
Many of the companies that signed the open weight AI letter have a direct economic stake in seeing open-weight models flourish. Nvidia sells GPUs. If hundreds of startups and enterprises run their own AI models, demand for those chips grows. AMD competes in the same market and signs for the same reason.
Cloud providers and infrastructure firms, from Microsoft Azure to Dell and Cisco, benefit when organizations deploy AI workloads on their own hardware or in hybrid clouds. Slashdot noted TechCrunch’s observation that “companies like Nvidia, Microsoft Azure, and other infrastructure providers have a vested interest in pushing for commoditized models: If models are interchangeable, people will buy more GPUs, rent more cloud capacity, and build more applications.”
Meta has put billions into releasing open-weight large language models like Llama. Making open models the default, not the exception, protects that investment. It also positions Meta’s technology as the foundation for a global developer ecosystem. Hugging Face, which hosts and distributes open models, signed too.
The open weight AI letter reads as both a policy statement and a market protection move. The united front sends a clear signal to regulators: the industry will resist efforts to lock down AI before a competitive market takes hold.
Startup Founders Fire Their Own Warning
A separate appeal from nearly 200 Silicon Valley founders, including people from Y Combinator and Proton, arrived around the same time as the open weight AI letter. That letter zeroes in on the risk of cutting off U.S. access to Chinese open-weight models like DeepSeek and Kimi.
The startup founders wrote that “American leadership requires two things: world-leading American open-weight models and continued access for U.S. builders to open models already available worldwide.” They argue that broad prohibitions would cripple the next generation of American startups that rely on readily available models to experiment and build products quickly.
Instead of blanket bans, the founders urged the government to adopt targeted safeguards. The fear is that restricting access would hand an advantage to competitors in other countries while doing little to improve security.
Both letters arrived as the White House accused China of using distillation to copy models from U.S. companies such as Anthropic. The timing shows how the distillation question sits at the center of the open-weight debate. It also explains why the open weight AI letter explicitly calls for protecting the technique through narrow legal frameworks, not sweeping prohibitions.
FAQ About the Open Weight AI Letter
Why did OpenAI sign the open weight AI letter if it sells closed models
OpenAI continues to offer both proprietary models and open-weight releases. Sam Altman said he wants the U.S. to win with both approaches. Greg Brockman added that he has “at a deep level” always believed that having more models and more usage is a good thing. OpenAI signing the open weight AI letter signals it can coexist with, and even benefit from, a diverse ecosystem.
What is model distillation and why does the letter protect it
Model distillation is a training method where a smaller AI learns by studying the outputs of a larger, more capable model. The technique does not copy the original model’s weights. It generates new training data from the larger model’s responses. The open weight AI letter argues that distillation is a normal engineering practice and should not be outlawed. It says intellectual property theft should be handled case by case, not by banning a widely used technique.
Which companies did not sign the open weight AI letter
Google and Anthropic stayed out. Both companies focus on closed, frontier models and have not released open-weight systems that rival their best proprietary offerings. Their absence highlights the split between firms that want to keep the most advanced capabilities behind paid APIs and those that want open access to fuel chip sales, cloud revenue, and developer ecosystems.
