Anthropic faces 2 government bans as OpenAI's 1B users draw scrutiny

Anthropic's legal fight with the Trump administration over defense access may prove less damaging than the structural political exposure OpenAI faces with its billion-user consumer base.

Anthropic is fighting two government actions — a Defense Department security-risk designation and Commerce Department export curbs — but the political headwinds battering OpenAI as it prepares for a potential $1 trillion IPO may prove more lasting.

"Anthropic's exclusion from defense contracts underscores the power of its models and their current leadership position," said Brian Pitz, an analyst at BMO Capital Markets.

The Defense Department designated Anthropic a supply-chain risk earlier this year after the company refused to grant unrestricted access to its tools for use in autonomous weapons and mass surveillance. Anthropic filed two lawsuits to block the designation. The Commerce Department separately used export-control authority to restrict foreign access to Anthropic's models, forcing the company to cut off its most-advanced Fable 5 and Mythos 5 systems. The administration later eased those restrictions after Anthropic addressed cybersecurity concerns.

The diverging regulatory trajectories carry implications for both companies as they pursue IPOs that could value each above $1 trillion. While Anthropic's corporate customers supply about 80% of its revenue, OpenAI's mass-market reach — more than 1 billion monthly active users as of May, according to Sensor Tower — exposes it to the kind of social-impact scrutiny that has dogged social-media platforms for years.

The Defense Department's action against Anthropic relied on statutes designed to prevent foreign adversaries from involvement in US defense procurement. The government appears likely to prevail in at least one of the two lawsuits Anthropic filed, allowing it to continue excluding the company from defense contracts. That will limit Anthropic's addressable market, though defense work was never a major revenue driver for the company.

The reputational calculus may cut both ways. Anthropic's refusal to allow its models in fully autonomous weapons or mass surveillance could strengthen its standing with corporate clients wary of being associated with military operations. App-download data supports that view: uninstalls of OpenAI's ChatGPT surged 295% early this year after news that OpenAI was negotiating its own Pentagon deal, according to Sensor Tower, while Anthropic's Claude surpassed ChatGPT in new installations during the same period.

OpenAI's Broader Exposure

OpenAI faces a different order of political risk. Chief Executive Sam Altman has already testified before Congress about AI risks and has proposed granting the US government a 5% equity stake in the company, a move that Kalshi prediction markets price at just 24% probability this year. The proposal, valued at about $42.6 billion based on OpenAI's $852 billion valuation, aims to preempt regulatory friction by giving Washington a financial interest in the company's success.

But government ownership carries its own hazards. A public-sector stake would subject OpenAI to political control that could hamper its competitive position, particularly as rivals including Anthropic have proposed alternative structures such as investment accounts for Americans affected by AI-related job displacement. The steady departure of safety-focused executives — including chief futurist Joshua Achiam, who is leaving after nearly nine years — adds to the governance concerns surrounding Altman's leadership.

Forward Scenarios

The regulatory environment for frontier AI remains volatile. Anthropic's Fable 5 model was released and then recalled at the administration's direction before export controls were later eased. OpenAI's GPT-5.6 rollout was similarly gated, with the Commerce Department clearing wider access only after additional testing and closed-door government meetings. The uneven access timeline shows how tightly frontier AI distribution is now tied to federal decisions.

For Anthropic, a resolution with the Defense Department appears likely when political conditions shift, given the Pentagon's incentive to access advanced AI tools. For OpenAI, the structural risks embedded in its consumer-facing business model — more than 1 billion users, congressional scrutiny, and the unresolved question of government equity — may prove more durable.

This article is for informational purposes only and does not constitute investment advice.