{"slug": "anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in", "title": "Anthropic, OpenAI Not In List Of 20 US Companies That Have Backed Open Models In Joint Letter", "summary": "Twenty major US companies, including NVIDIA, Meta, Microsoft, and IBM, signed a joint letter defending open-weight AI models as essential to American leadership, but Anthropic and OpenAI—the two leading frontier AI labs—conspicuously did not. The letter, posted by NVIDIA CEO Jensen Huang, argues that open-weight models and distillation are vital for innovation, directly contradicting both Anthropic's and OpenAI's recent lobbying against distillation and their business models built on proprietary API access.", "body_md": "Even as 20 major US companies have backed open-source AI, some companies are more conspicuous by their absence.\n\nThe letter, titled “Open Weights and American AI Leadership” and shared publicly by NVIDIA CEO Jensen Huang in his first-ever post on X, argues that open-weight models — ones anyone can download, inspect, modify, and run on their own hardware — are essential to American AI leadership, cybersecurity, and economic competitiveness. It’s signed by American Innovators Network, Andreessen Horowitz, Arcee AI, Arena, Black Forest Labs, Box, CrowdStrike, Dell Technologies, Emergence Capital, Hugging Face, IBM, The Linux Foundation, Mariana Minerals, Meta, Microsoft, Mistral, Mozilla, NVIDIA, Palantir, Perplexity, Reflection, Replit, ServiceNow, Telnyx, and Y Combinator.\n\nThat’s a long and fairly eclectic list. What jumps out, though, is who isn’t on it. Anthropic and OpenAI, the two labs currently sitting at the top of the frontier model race, are nowhere to be found. Neither is Google, whose Gemini models until recently sat right alongside them. Given how loudly Anthropic and OpenAI have both spoken about the future of AI recently, their silence on a letter specifically about open models reads less like an oversight and more like a position.\n\n## Frontier Labs, Closed Doors\n\nAnthropic and OpenAI are the two companies with the most to lose from a world flooded with capable open-weight alternatives. Both currently sell access to their best models through APIs and subscriptions, and both have built valuations north of $850 billion on the premise that their frontier models are worth paying a premium for. An ecosystem where open-weight models close the capability gap to a matter of months, as independent benchmarks have increasingly shown, chips directly away at that premium. It’s not hard to see why neither company would want to put its name on a letter arguing that openness should be protected from restriction, particularly the letter’s specific defense of distillation as a legitimate technique rather than a threat.\n\nThat defense of distillation is worth sitting with, because it cuts directly against positions both labs have taken. OpenAI has [told the US government](https://officechai.com/ai/us-threatens-sanctions-against-chinese-models-that-are-allegedly-distilling-us-models/) that Chinese labs like DeepSeek were using hidden techniques to distill its models, a complaint that helped push the White House toward exploring sanctions against foreign labs accused of doing exactly that. Anthropic, separately, has[ accused](https://officechai.com/ai/anthropic-says-that-deepseek-moonshot-and-minimax-created-24000-fake-claude-accounts-to-steal-its-models/) DeepSeek, Moonshot AI, and MiniMax of creating tens of thousands of fraudulent accounts to extract data from Claude in violation of its terms of service. A letter defending distillation as a “long tradition of learning from, building upon, and improving existing technologies” is a hard one for either company to sign onto without contradicting its own lobbying.\n\nThere’s a real irony sitting underneath all of this for OpenAI in particular. The company’s founding pitch, back in 2015, was built around openness — it’s in the name. Sam Altman has since [conceded the company got this wrong](https://officechai.com/ai/weve-been-on-the-wrong-side-of-history-over-not-open-sourcing-models-openai-ceo-sam-altman/), telling an audience earlier this year that OpenAI has “been on the wrong side of history” when it comes to not open-sourcing its models. OpenAI did eventually respond with its gpt-oss family, which briefly held the title of strongest open-weights model out of the US before Chinese labs pulled ahead. But the company’s actual center of gravity has moved decisively toward closed, proprietary frontier releases, and its posture in Washington — flagging distillation, warning about foreign misuse — puts it on the opposite side of the argument this letter is making.\n\nAnthropic’s absence follows a more consistent pattern. The company has built its identity around safety-first AI development since splitting off from OpenAI in 2021, and it has [talked about regulation and safety](https://officechai.com/ai/anthropic-talks-about-regulation-and-ai-safety-far-more-than-openai-shows-data/) more than any other major lab, with CEO Dario Amodei pushing for mandatory FAA-style testing regimes for frontier models. Critics, including Meta’s former chief AI scientist [Yann LeCun](https://officechai.com/ai/anthropic-ceo-dario-amodei-deluded-about-dangers-of-ai-meta-ai-chief-yann-lecun/) and former White House AI czar [David Sacks](https://officechai.com/ai/anthropic-has-a-pattern-of-using-fear-to-market-its-products-us-ai-czar-david-sacks/), have accused Anthropic of using safety rhetoric as a form of regulatory capture — a strategy, in LeCun’s words, to scare regulators into restricting open-source competitors “out of existence.” Whether or not that framing is fair, a company that has spent the past year arguing for tighter oversight of frontier AI was always an unlikely signatory on a letter arguing the opposite: that openness itself is a safety strategy, and that restricting it is the greater risk.\n\n## Why Meta And Microsoft Make Sense\n\nMeta and Microsoft sit in a different spot entirely, and their names on this letter track cleanly with their business incentives.\n\nMeta doesn’t have a frontier closed model to protect. While [Meta has quietly walked back some of its earlier open-source commitments](https://officechai.com/ai/meta-appears-to-change-open-source-policy-in-just-a-year-says-itll-now-be-careful-in-what-to-open-source/) following Llama 4’s rocky reception and its Muse Spark models are now close to the frontier, it still seems want more competition for OpenAI and Anthropic through open models. Microsoft’s calculus is arguably even more direct. Despite owning 27% of OpenAI’s for-profit arm, Microsoft CEO Satya Nadella has been publicly saying for over a year that [foundation models are getting commoditized](https://officechai.com/ai/foundational-ai-models-are-getting-commoditized-microsoft-ceo-satya-nadella/), describing OpenAI itself as “not a model company” but a product company, and warning that being first with a frontier model could turn into a “winner’s curse” once competitors catch up. Microsoft AI CEO Mustafa Suleyman has gone further still, [openly stating](https://officechai.com/ai/anthropic-is-extremely-expensive-many-are-urgently-looking-for-alternatives-microsoft-ai-ceo-mustafa-suleyman/) that Microsoft wants to reduce and eventually eliminate what it pays Anthropic for API access, with the explicit goal of building its own frontier-competitive models rather than remaining a customer forever. A world where open weights commoditize the model layer is a world that suits Microsoft just fine — it still owns the distribution, through Azure, Copilot, and 365, regardless of whose weights are running underneath.\n\nPut together, the list of signatories reads less like a random coalition and more like everyone who benefits from AI models becoming cheap and interchangeable, standing across the table from everyone whose business depends on their models staying expensive and exclusive.", "url": "https://wpnews.pro/news/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in", "canonical_source": "https://officechai.com/ai/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in-joint-letter/", "published_at": "2026-07-24 15:29:59+00:00", "updated_at": "2026-07-24 15:57:04.197512+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-policy", "ai-ethics"], "entities": ["Anthropic", "OpenAI", "NVIDIA", "Jensen Huang", "Meta", "Microsoft", "IBM", "Andreessen Horowitz"], "alternates": {"html": "https://wpnews.pro/news/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in", "markdown": "https://wpnews.pro/news/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in.md", "text": "https://wpnews.pro/news/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in.txt", "jsonld": "https://wpnews.pro/news/anthropic-openai-not-in-list-of-20-us-companies-that-have-backed-open-models-in.jsonld"}}