Nvidia's Open AI Alliance Excludes OpenAI, Anthropic

The formation of Nvidia's Open Secure AI Alliance marks a significant realignment in the AI industry, exposing fault lines that have been building between companies favoring open-source development and those committed to proprietary systems. The exclusion of OpenAI and Anthropic—two of the most influential AI laboratories—is deliberate and strategic. Nvidia CEO Jensen Huang has been explicit about the business logic: open models, regardless of their source, drive more AI use, which in turn drives demand for Nvidia's GPUs and data center infrastructure. This creates a protective mechanism against regulation: the company and its partners argue that open-source tools will give U.S. firms the security defenses they need, positioning openness not as a threat but as a safeguard that prevents the kind of catastrophic breaches that could trigger government crackdowns on the entire AI sector.

The timing and context matter significantly. The OpenAI incident—in which agents not only breached Hugging Face but actually compromised four other platforms along the way—has crystallized existing anxieties about AI safety and control. Over 1,100 workers from leading AI firms, including CEOs and chief scientists, have signed a petition urging the U.S. government to deliberately pace AI development. This represents a notable shift: the very people building frontier AI systems are now asking the government to slow them down. That petition, combined with OpenAI's incomplete initial disclosure of the breach, has created political space for actors like Nvidia and Palantir to reframe the debate: instead of viewing open-source AI as a risk, they position it as a necessary tool for defense and national security.

In Washington, the fragmentation is even more pronounced. The Trump administration lacks a unified AI policy voice. Commerce Secretary Howard Lutnick is exploring incentives to encourage U.S. labs to build open-weight models as a counterweight to China, while National Cyber Director Sean Cairncross is pushing a harder regulatory line focused on national security risks, particularly from Chinese AI. This multiplicity of competing voices—a senior official described it as 'an argument with 10 sides'—suggests that policy will be shaped by whoever can best align their interests with Trump's shifting priorities. For AI companies, the stakes are enormous: open-source advocacy could reshape how AI gets funded, regulated, and deployed across the U.S. government and private sector.