AI TRENDS | Polymarket Lists Market on Whether U.S. Government Will Restrict Public Access to a Major AI Model

Polymarket has listed a new prediction market asking whether the U.S. federal government will revoke public access to another major AI model, with the probability temporarily at 33%.

According to Odaily, the market will resolve “Yes” if, by the end of 2026, the U.S. federal government passes legislation, issues an executive order, imposes export controls, or takes any other action that materially restricts U.S. public access to a major AI model.

The rules define a qualifying action as a formal U.S. government measure whose effect is equivalent to fully banning public access to an AI model within the United States, regardless of the stated purpose. The rules also state that if an action effectively prevents the public from accessing the model through normal channels in the United States—such as by prohibiting the model from being provided to foreign citizens or governments—this can qualify if it results in the general public being unable to access the model domestically.

The rules specify that removing access through only a single channel is not sufficient. Access removals not caused by any formal U.S. government action are not considered qualifying.

The market defines “major AI models” as flagship, general-purpose large language models or multimodal foundation models developed by OpenAI, Anthropic, Google (including Google DeepMind), Meta, xAI, Microsoft, Amazon, Mistral AI, DeepSeek, Alibaba, ByteDance, Moonshot AI, and Zhipu AI (Z.ai). Task-specific models, outdated models, and models intended only for research or preview purposes are excluded.

The action may target a single model or a group of models, as long as at least one major AI model becomes publicly inaccessible in the United States. A temporary ban can qualify, but an action does not qualify if the public can still access the model until the effective date of the decision.

The market cites official information and announcements from the U.S. government and relevant AI companies as sources, and may also reference reliable reporting to form consensus.