Will the US Government Ban Another Major AI Model in 2026?

The US government already pulled the trigger once. In June 2026, the Commerce Department used national security export controls to force Anthropic to suspend all public access to Claude Mythos 5 and Fable 5, its two most capable frontier models. The ban lasted roughly two weeks before Washington reversed course. Now the question hanging over every AI lab in the country is whether the government does it again. The market puts the odds at 43.5 percent that a second major AI model loses public access before December 31.

This contract resolves YES if the US government removes public access to another major AI model in 2026, beyond the Anthropic episode already on record. The YES outcome sits at 43.5 percent; the NO outcome sits at 56.5 percent. The market resolves December 31, 2026, and has seen $5,322 in total lifetime volume, all of it recorded in the last 24 hours, which tells you this contract is fresh and still finding its floor.

How the US Government AI Access Contract Works

The YES outcome pays if Washington formally directs a company to suspend or remove public access to a major AI model during calendar year 2026, excluding the Anthropic Mythos 5 and Fable 5 action already resolved. A new executive order, a Commerce Department export control directive, or a court-ordered shutdown targeting a different model would each qualify. The resolution body is Polymarket’s own market resolution process.

• YES outcome (43.5 percent): The US government restricts public access to at least one additional major AI model before December 31, 2026.
• NO outcome (56.5 percent): No further government-mandated removal of public access to a major AI model occurs in 2026.

The NO outcome pays out if the Anthropic episode remains the only instance of government-compelled access removal in 2026. That means OpenAI, Google DeepMind, Meta, xAI, and every other frontier lab finish the year with their flagship models publicly available. Given the Commerce Department’s willingness to act against Anthropic and the bipartisan pressure building around AI national security, that outcome is far from guaranteed, but the market still prices it as the more likely result.

Market Signals Point to Genuine Uncertainty

The momentum composite here is striking. The 1-hour price change is flat at zero, but the trend score registers 14, which is elevated and signals that directional conviction is building beneath a quiet surface. The Anthropic ban lifted June 30 and the lift generated immediate market activity, which explains why the full 24-hour volume of $5,322 is also the lifetime total. This contract is essentially one day old, and the trend score tells you traders are watching closely.

Liquidity sits at $8,367 against $5,322 in volume, meaning order-book depth exceeds traded volume. That ratio is healthy for a new market and suggests market makers are prepared for a price move. Open interest is currently zero, which is consistent with a market that opened and traded but has not yet accumulated a long-duration position base. The math here is straightforward: genuine two-sided uncertainty, not a market drifting toward consensus.

• The Commerce Department used export controls against Anthropic on June 12, established that the legal mechanism exists and is ready to deploy again.
• The ban on Mythos 5 and Fable 5 lasted roughly 18 days, showing Washington can reverse quickly, but the precedent of acting at all is now set.
• The Trump administration reversed the Anthropic controls after criticism from tech investors and executives, suggesting political pressure can move policy both directions.
• OpenAI, Google DeepMind, and Meta each operate frontier models that could attract national security scrutiny under the same legal framework used against Anthropic.
• The momentum composite, flat price movement combined with a trend score of 14, reflects a market digesting the Anthropic reversal while watching for the next catalyst.

Lines Analysis: What Moves This Market Before December

The 43.5 percent YES probability reflects the real precedent set in June. The Commerce Department demonstrated both the willingness and the legal authority to compel access removal. Any new national security finding targeting a Chinese-linked model, a foreign-national access concern at another lab, or an escalation in the US-China AI competition could trigger a second action before year end. The administration has roughly six months left in 2026 to act, and the regulatory appetite appears intact despite the Anthropic reversal.

The NO outcome at 56.5 percent is not a statement of safety. The NO outcome depends on the administration choosing restraint after the Anthropic episode generated significant backlash from the tech community and handed non-American competitors a brief visibility advantage. Commerce Secretary Howard Lutnick personally coordinated the Anthropic reversal, signaling that the administration is sensitive to blowback. A second action would face even louder opposition, raising the political cost of YES.

• A new executive order expanding national security review of frontier AI models would push YES probability sharply higher.
• Congressional hearings on AI export controls in the second half of 2026 could either accelerate or restrain a second government action.
• Any public incident linking a major AI model to foreign intelligence activity would move this market fast toward YES.
• The Commerce Department’s track record of reversal on the Anthropic case is a moderating signal: acting only to reverse quickly raises the reputational cost of a repeat.
• Meta’s open-source Llama models and xAI’s Grok represent different legal challenges for access removal, since open-source distribution complicates export control enforcement and could push the market in either direction depending on how Washington interprets its own legal authority.

Lifetime volume of $5,322 is low by any standard. The data does not yet favor either side with the conviction a high-volume market would provide. What the data does show is a market priced at genuine 44-to-56 uncertainty, not a market drifting toward a conclusion. Six months remain on the calendar, and the administration has shown it will act when it decides the national security case is clear enough.

LINES VERDICT

Lean NO, but watch the calendar

The Anthropic reversal raised the political cost of a second ban, and the administration has shown sensitivity to tech-sector pushback. Without a new triggering event, the government is unlikely to repeat the same move so quickly.

What the market says: The market prices this at 43.5 percent YES, treating a second government-mandated access removal as a real but minority outcome. With six months remaining and low total volume, this probability will move fast the moment a new national security action surfaces.

Political Context

The Anthropic Claude Mythos 5 and Fable 5 episode is the clearest base rate available for this market. The Commerce Department acted on June 12, 2026, citing national security export control authority. Access was suspended for all customers, not just foreign nationals, because Anthropic could not reliably verify user nationality at scale. The restriction generated immediate criticism from technology executives and investors, partly on the grounds that the ban handed foreign competitors a window of advantage. The Commerce Department lifted the controls by June 30, 2026, after coordinating directly with Anthropic on a compliance framework. That 18-day cycle is the only direct precedent the market has to price against, and it cuts both ways: the government acted, and the government reversed. Both facts matter for where the 43.5 percent probability sits today.

Related Prediction Markets

• Politics Prediction Markets Hub: track all active US government and regulatory contracts in one place.
• Will the US Government Issue New AI Export Controls in 2026? A directly related question covering the broader regulatory environment driving this contract.
• US-China AI Policy Escalation in 2026: the geopolitical backdrop most likely to trigger a second government action in the AI access space.