Nvidia, Microsoft and allies urge U.S. to back open-weight AI models

Welcome to AI Decoded, Fast Company’s weekly newsletter that breaks down the most important news in the world of AI. I’m Mark Sullivan, a senior writer at Fast Company, covering emerging tech, AI, and tech policy. Sign up to receive this newsletter every week via email here. And if you have comments on this issue and/or ideas for future ones, drop me a line at sullivan@fastcompany.com, and follow me on X @thesullivan.  The AI industry rallies around open models. Is it more than talk? Most of the big AI model makers keep the parameter weights, the billions of settings that work together to produce useful outputs, of their best models hidden from developers and other users. But this week, the AI industry has been falling all over itself to embrace models that expose those weights. On July 24, a coalition led by Nvidia, Microsoft, and Meta published “Open Weights and American AI Leadership,” a three-page letter asking Washington to avoid premature restrictions on downloadable models. Jensen Huang promoted it in his first post on X, and the list of signatories doubled to 50 within a day, adding OpenAI, Google, AMD, Cisco, and GitHub. Only Anthropic chose not to sign. The company’s CEO, Dario Amodei, wrote in a blog post Monday that Anthropic has never sought to ban open-weights models. But he argued that continuing to restrict the most powerful chips, most of which come from Nvidia, in places such as China is the right way to limit the creation of potentially dangerous AI models. He also advocated for the government to punish foreign model makers that use the outputs of Western AI models for large-scale training. Mark Zuckerberg took the open-weights argument to The Wall Street Journal on Tuesday, writing that superintelligent models are coming and that the defining question of our time is who gets access to them. Zuckerberg argues that it is better to distribute open models widely so developers can help make them safer. He also contends that allowing a few big tech companies, working closely with the government, to control powerful closed models is ultimately more dangerous. The irony is that Meta has largely given up on open models with the demise of its Llama line. The company is now placing its bets on massive closed models built by its new AI organization, Meta Superintelligence Labs, led by Alexandr Wang. Zuckerberg’s op-ed is being lauded across the AI industry, but it is also a perfect illustration of how cheap advocacy and good intentions can be. Meta’s new Muse Spark models are unlikely to become open-weights anytime soon. The economics of developing frontier AI models make closed weights increasingly inevitable. The work is hugely expensive, and investors such as Andreessen Horowitz, SoftBank, and Amazon, which are placing big bets on closed AI labs, will not earn the returns they want if the product is free and open to everyone. They want an AI market in which enterprises and startups depend on a handful of major providers that charge a premium for access to closed models. Until entirely new business models emerge that push Silicon Valley capital toward the development and distribution of open models, I will remain skeptical of the open-weights rah-rah from people such as Zuckerberg, Jensen Huang, and former Trump AI adviser David Sacks. For now, developers at enterprises and AI startups will continue choosing Chinese open-weights models such as Kimi 3 and DeepSeek V4 for most AI work, while reserving closed U.S. models from OpenAI and Anthropic for the hardest jobs. More than 1,100 AI workers call on Washington to ‘pace’ AI progress Bloomberg reports that more than 1,100 employees across nearly a dozen AI companies, including OpenAI, Anthropic, Google, and Meta, signed a petition urging the U.S. government to create a mechanism to “deliberately pace” AI development. The letter, titled “Pacing the Frontier,” was signed by Anthropic CEO Dario Amodei, OpenAI chief scientist Jakub Pachocki, Meta chief scientist Shengjia Zhao, and Google AI safety head Anca Dragan. It comes after OpenAI disclosed that two test models escaped a lab environment, reached the open internet, and breached another company’s internal system. Why Google DeepMind employees are depressed about slow progress in AI coding Axios reported last week that poor morale at Google DeepMind is contributing to delayed model releases, with Gemini 3.5 Pro reportedly several months behind schedule. One employee told the outlet that DeepMind has been slow to improve its models’ AI coding capabilities, a claim that a recently departed DeepMind researcher confirmed to Fast Company. The stakes extend beyond coding as a product category. AI coding agents increasingly help develop and improve the models themselves, and these tools have noticeably accelerated model releases at OpenAI and Anthropic. Google DeepMind cannot afford to fall further behind. Keeping pace will require a first-class AI coding tool. Nvidia may backstop a $500 billion OpenAI data center  Nvidia is reportedly in talks to provide a financing guarantee of as much as $250 billion to help OpenAI lease a $500 billion, 10-gigawatt computing hub that SoftBank is developing in Ohio, with the site targeted for 2028. The backstop would let OpenAI raise construction and lease debt on Nvidia’s credit, which matters because OpenAI is unprofitable and cannot obtain an investment-grade rating on its own.  AI investors’ chilly summer continues   Investors continue to question how much Big Tech is spending on AI infrastructure. The impact is being felt mainly in the chip sector, which had seen strong growth much of this year. The Nasdaq 100 moved toward a correction Tuesday as the semiconductor selloff deepened. After a record high in late June, the Philadelphia Semiconductor Index had fallen 20% from that peak by Friday. Korea’s KOSPI index, which is dominated by memory chipmakers, had an up-and-down week last week and lost 5.72% of its value Friday. Pressure is building on the largest AI spenders to justify their capital budgets. Nvidia announced a major investment (reportedly $5 billion) in Ilya Stutskever’s Safe Superintelligence (SSI) Safe Superintelligence, the lab founded by former OpenAI chief scientist Ilya Sutskever, announced on Monday a long-term partnership with Nvidia that includes an undisclosed investment (Bloomberg reports $5 billion) and access to Nvidia’s Vera Rubin platform. The agreement is expected to increase SSI’s compute resources by an order of magnitude. Bloomberg reported the equity commitment at $5 billion, one of the chipmaker’s largest funding deals of the AI boom. SSI has been a remarkably secretive company: You’ll find no leaks to the press about its operations, and it’s published no research or shipped any products since launching in 2024. SSI was last valued at $32 billion. More AI coverage from Fast Company:  Claude users’ shared conversations were showing up in Google searches Meet the new robot dog patrolling LaGuardia Airport Should AI companies be able to outsource safety? OpenAI tells ChatGPT to stop impersonating famous authors Want exclusive reporting and trend analysis on technology, business innovation, future of work, and design? Sign up for Fast Company Premium.

The letter—signed by companies including Microsoft, Nvidia, OpenAI, SpaceX, Google, IBM, Palantir, and Meta—stressed the importance of American AI leadership

Leading technology firms advocate for open-weight AI models to bolster American innovation. This approach expands access and fosters competition across various industries. Open models offer organizations greater control over their data and AI capabilities. Transparency in these models enhances security and allows for broader vulnerability identification. Embracing open weights is seen as crucial for future economic prosperity and technological leadership.

With China's open-weight AI models continuing to improve, US companies want to see a free market for all.

A group of American tech giants, including Nvidia, Palantir and Meta, signed a public letter on Friday calling for US leadership in open-weight AI models, an area that China currently dominates. The challenge comes amid growing speculation that the Trump administration might introduce restrictions on open AI models. These would aim to address national security risks of the fast-moving technology and combat the growing popularity of Chinese models, a move that critics say would instead hurt US...

Nvidia, Microsoft, Meta, Palantir, and more than 20 other tech companies signed an open letter urging policymakers not to impose "premature restrictions" on open-weight AI models, warning that broad limits could "stifle competition or drive innovation overseas." CNBC reports: They wrote that open-weight models strengthen competition and ensure that the benefits of the technology are "broadly shared rather than concentrated in a few hands." "Relying solely on closed models is not inherently safe: they can be breached, misused, or fail in ways that outsiders cannot detect," the letter said. "And concentrating advanced AI capabilities behind a small number of closed models compounds that risk." Elon Musk, who runs an AI business under his rocket company SpaceX, also applified the letter on social media, writing that it has his "full support" in a post on X. SpaceX did not officially sign the letter. Greg Brockman, OpenAI's president, said Thursday that the company believes in broad access, and that he has not been involved in any conversations with the Trump administration about potentially banning Chinese open-weight models in the U.S. "I think that, that fundamentally, AI and AI usage is something that is actually very important to democratize," Brockman told reporters during a briefing in New York City. "And so, for me, at a sort of deep level, I think that having more models, more usage, that is a good thing." OpenAI CEO Sam Altman addressed the letter in a post on X on Friday, writing that he wants the U.S. to win with both open-weight and proprietary models, and that he is "glad to see this." [...] In the letter on Friday, the U.S. tech companies said that concerns about unlawful distillation should be addressed through "targeted legal and commercial frameworks" instead of with "sweeping restrictions on techniques that play an important role in AI innovation." "Our AI leadership will be judged not by one frontier AI model, but by whether the United States builds a strong, open ecosystem that diffuses into every sector," the letter said. "This is essential for creating opportunities for innovation and prosperity across the country." The letter follows a separate appeal signed by nearly 200 Silicon Valley companies, including Proton and Y Combinator, warning that restricting U.S. access to Chinese open-weight AI models could cripple the next generation of American startups. "American leadership requires two things: world-leading American open-weight models and continued access for U.S. builders to open models already available worldwide," the startup founders wrote. Instead of broad prohibitions, they argue the government should adopt targeted safeguards. Of course, these signees "have an obvious economic stake in seeing open AI models flourish," notes TechCrunch. "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." Read more of this story at Slashdot.