Resolution criteria

This market resolves YES if any country officially bans or heavily restricts the use of advanced AI systems before January 1, 2027. Resolution is based on official government announcements or legislation that explicitly prohibits or substantially restricts deployment of advanced AI systems (such as large language models, foundation models, or general-purpose AI systems).

A "ban" means prohibition of use or deployment. A "heavy restriction" means requirements that effectively prevent widespread commercial use, such as: mandatory licensing with severe limitations, operational bans in key sectors, or restrictions requiring systems to be substantially modified or disabled.

Restrictions on specific AI applications (e.g., biometric surveillance, emotion recognition) do not count unless they effectively prevent the use of advanced AI systems broadly. Restrictions on imports, exports, or access to AI chips do not count. Restrictions limited to government agencies or specific sectors do not count unless they represent a de facto ban on the technology.

Resolution sources: Official government websites, legislative databases, and major news outlets reporting on official policy announcements.

Background

The EU AI Act officially became law on August 1, 2024, with implementation staggered from early 2025 onwards. 'Unacceptable-risk' bans and AI literacy requirements have applied since February 2, 2025, while general-purpose AI rules began applying August 2, 2025. However, high-impact general-purpose AI models that might pose systemic risk, such as the more advanced AI model GPT-4, would have to undergo thorough evaluations and any serious incidents would have to be reported to the European Commission—this is regulation rather than a ban.

As of 2025, ChatGPT is restricted in over two dozen countries. However, these restrictions vary widely in scope and nature. Not all countries on the list have active bans. In places like Hong Kong and Belarus, the service is unavailable because OpenAI does not support accounts there. These exclusions are often due to business, regulatory, or legal complexities rather than explicit censorship.

China does not yet have a standalone, comprehensive AI statute. The GenAI Interim Measures, which came into force in August 2023, were the first comprehensive generative AI regulations in China, governing the development, governance, service provision, supervision and use of generative AI systems. China's strategically lenient approach to regulation may therefore offer its AI firms a short-term competitive advantage over their European and U.S. counterparts.

Considerations

The distinction between regulation and restriction is critical. Most major jurisdictions are implementing regulatory frameworks rather than bans. The EU AI Act imposes compliance requirements on advanced AI systems but does not ban them. China regulates generative AI but has not banned advanced systems. The US has not implemented comprehensive AI legislation at the federal level.

A country would need to take an unusually restrictive stance—effectively prohibiting the deployment or use of advanced AI systems—for this market to resolve YES. Regulatory requirements, licensing schemes, or restrictions on specific applications do not constitute a ban or heavy restriction under these criteria.