This prediction market asks whether the United States will enact federal legislation addressing artificial intelligence safety by the end of 2026. Specifically, it resolves to 'Yes' if a bill signed into law contains at least one of two types of provisions. The first is a prohibition on the creation or release of specific AI systems or models. The second involves setting limits on AI training, such as restricting access to certain data or imposing a maximum limit on the number of parameters used in model training. The question reflects growing political and public concern about the potential risks of advanced AI, including national security threats, economic disruption, and loss of human control over autonomous systems. Legislative activity on AI has increased in the 118th Congress, with multiple committees holding hearings and drafting framework bills. However, significant disagreements exist between lawmakers, the tech industry, and civil society groups on the appropriate scope and stringency of regulation. The 2026 deadline accounts for the current congressional session and the next, accounting for typical legislative timelines in an election year. Interest in this market stems from uncertainty about whether political consensus can be reached on what would constitute a major new regulatory framework for a foundational technology.

Historical Context

Federal attempts to regulate emerging technologies provide a template for AI legislation. The 1996 Communications Decency Act and subsequent laws attempted to govern the early internet, often playing catch-up with technological change. More recently, the 2022 CHIPS and Science Act included provisions for AI research but did not establish safety regulations. The direct precedent for AI safety legislation is sparse. In 2023, the European Union finalized its AI Act, which takes a risk-based approach and includes some prohibitions on certain AI uses. This marked the first major comprehensive AI law by a Western democracy and has influenced the U.S. debate. Domestically, the primary federal action has been executive. President Biden issued Executive Order 14110 on 'Safe, Secure, and Trustworthy Artificial Intelligence' in October 2023. This order used existing agency authority to mandate safety testing for powerful AI models and develop standards, but it did not create new prohibitions or training limits that require congressional action. Several states, including California and Colorado, have begun passing their own AI laws, creating a potential patchwork that increases pressure for a federal standard. Past congressional efforts on tech regulation, such as privacy bills, have frequently stalled due to partisan divides and industry lobbying, suggesting a high barrier for passage of restrictive AI safety measures.

Frequently Asked Questions