The Trump administration has moved to restrict access to advanced artificial intelligence models developed by OpenAI and Anthropic, companies actively pursuing public offerings on Wall Street, demonstrating the state's direct intervention in the control and deployment of emerging technologies for capital. OpenAI announced Friday it is restricting the release of its new AI model, GPT-5.6 Sol, at the request of President Donald Trump’s administration. This restriction means the product will only be accessible to customers approved by the Trump administration, a step OpenAI described as temporary on the “path to broader availability in the coming weeks.”
Anthropic, a chief rival to OpenAI, followed hours later with its own announcement that the Trump administration had approved a limited release of its strongest cybersecurity model, Mythos 5. This approval came two weeks after the U.S. Commerce Department had effectively banned the model. Both companies stated their newest models would be available only to small groups of trusted partners, indicating a controlled distribution that prioritizes specific corporate or state-aligned entities.
The State's Hand in Capital
The government’s heightened AI oversight follows actions taken by the Commerce Department earlier this month against Anthropic, which led the company to take offline two new AI models, Fable 5 and Mythos 5, just days after their unveiling. These models were blocked from use by foreign nationals under a Trump directive. Officials cited concerns raised by Anthropic itself earlier this year, when the company warned that its Mythos model was adept at finding software flaws that could be weaponized by malicious hackers, potentially threatening critical computer networks globally.
President Trump signed an executive order on AI oversight earlier in June, establishing a framework for the federal government to vet the national security risks of advanced AI systems for up to 30 days before public release. While described as voluntary, this framework has not yet been fully developed, leaving the process opaque. The Pentagon previously designated Anthropic as a national security risk for raising ethical and safety concerns about AI usage in war, and Trump himself ordered federal agencies to stop using Anthropic’s Claude chatbot. Anthropic responded to these state actions with a lawsuit that remains in federal courts.
Managing Risk for Profit
The government’s intervention adds a new layer of complexity to the exploratory moves by both OpenAI and Anthropic to take their companies public on Wall Street. These companies are following the path of SpaceX, which completed a record-setting initial public offering 15 days ago. President Trump has publicly entertained the idea of the U.S. government owning a stake in leading AI companies, proposing a concept where “pieces could be given to the American public, where the American public essentially becomes a partner with the companies.” This proposition suggests a form of state-capitalist control, potentially consolidating wealth and power under a different guise.
Investor David Sacks, a co-leader of Trump’s council of technology and science advisers, publicly attributed the heightened government scrutiny to Anthropic CEO Dario Amodei. Sacks stated on a recent podcast that Amodei “spiked the cortisol level, got everyone really worried” by claiming to have created a “cyber weapon” called Mythos, acknowledging “some truth to it in terms of the sense that this model had advanced cyber capabilities.” This narrative frames corporate innovation as a national security threat, justifying state control over its deployment.
Liberal Solutions and Systemic Limits
U.S. Rep. Lori Trahan, a Massachusetts Democrat and co-author of a bipartisan bill to regulate AI, criticized the administration’s approach, stating her concern that “the Trump administration is deciding company by company who gets access to the newest AI model. No law. No process. No oversight. Just appointees in Washington deciding who’s in and who’s out.” Trahan’s critique focuses on procedural fairness within the existing system rather than challenging the fundamental private ownership and state control of such powerful technologies. Stanford University cybersecurity expert Alex Stamos, chief product officer at AI security company Corridor, also expressed skepticism, stating that “pretty much nobody in the cybersecurity industry believes that there’s any factual basis for this action.” Stamos, a former chief security officer at Facebook parent Meta, reviewed an analysis of Anthropic’s Fable model and found no unique risks compared to other publicly available AI models, including those from China. He argued that such restrictions hinder the United States in the global technological race, revealing the underlying imperial competition for dominance.
OpenAI CEO Sam Altman and Anthropic executives have engaged in a series of negotiations with Trump officials, including U.S. Commerce Secretary Howard Lutnick, in recent weeks. Anthropic expressed its satisfaction with the partial release of Mythos and affirmed its commitment to “continue to work with the government to expand access” and make Fable available again to general users, indicating a corporate willingness to align with state directives for market access. Lutnick acknowledged Anthropic’s work in addressing government concerns, stating it “yielded significant progress,” solidifying the collaboration between state power and private capital in managing advanced technology. The state, through its regulatory and enforcement arms, continues to shape the landscape for capital accumulation in the burgeoning AI sector, ensuring that access and deployment serve established interests. The public, meanwhile, remains largely excluded from these critical developments, awaiting the limited and vetted releases approved by both corporate and state actors. The state’s actions here are not a flaw in the system, but rather a demonstration of its function: to protect accumulated wealth and manage the contradictions of advanced technology in service of capital. The proposed reforms, such as those by Rep. Trahan, seek to manage the system's contradictions without addressing its foundations, extending its life rather than challenging its core structures of power and ownership. The ongoing negotiations between AI executives and Trump officials underscore the intimate relationship between state power and corporate interests, ensuring that the development and deployment of transformative technologies remain under the control of a select few, rather than serving the collective good. The public remains an afterthought, offered symbolic concessions like the vague promise of "pieces" of ownership, while the real power dynamics continue to concentrate wealth and control upwards. The state's role as an enforcer for capital is clearly demonstrated in these actions, shaping the future of AI to serve existing power structures rather than challenging them. The restrictions on these powerful AI models, ostensibly for national security, ultimately serve to consolidate control over a critical emerging technology within the hands of the state and its favored corporate partners, further entrenching the existing economic order.