A proposal to establish a People’s AI Constitution Council in the United States aims to introduce democratic oversight into the governance of transformative artificial intelligence by conditioning federal contract eligibility on AI models adhering to constitutions selected by a randomly chosen citizen body. This approach seeks to remedy the current lack of public involvement in shaping the values embedded in frontier AI technologies deployed for government use.
What Happened
The proposal outlines the creation of a Council composed of 100 randomly selected U.S. citizens, each serving two-year terms, tasked with selecting three model AI constitutions from submissions by leading AI labs, research organizations, and the public. The Council would choose, not draft, these constitutions, which must meet rigor criteria verified by the Center for AI Standards and Innovation (CAISI). Adoption of one of these constitutions would become a condition for AI models to qualify for federal contracts. The Council’s selections would be reviewed biannually, with CAISI able to recommend amendments following technological advances, ensuring ongoing suitability and risk mitigation.
Key Facts
- Jurisdiction: United States federal government
- The People’s AI Constitution Council would consist of 100 citizens, randomly selected nationwide.
- Council members receive a three-month AI and constitutional AI training before serving.
- CAISI evaluates submitted constitutions and monitors AI model compliance but does not make constitution selections.
- AI model developers are not legally required to adopt any constitution but risk exclusion from federal contracts if they do not.
- Submissions to the Council could come from frontier AI labs, nonprofit research organizations, university programs, and public groups with at least 100,000 electronic signatures.
- Constitutions remain valid for two years, subject to periodic review and potential amendments.
- The measure has not been legislated or enacted and currently remains a proposal.
Why It Matters
This proposal addresses the democratic deficit in AI governance by empowering ordinary citizens to influence the foundational values guiding transformative AI systems used by government agencies. It aims to supplement technical alignment efforts led by private AI labs with a form of social legitimacy and oversight through public participation, a factor currently missing from AI model development. Conditioning federal AI procurement on adherence to citizen-selected constitutions may create cascading incentives across the AI ecosystem, potentially influencing broad adoption of accountability standards.
Background
The digital governance framework lacks established democratic controls over AI model values, traditionally shaped solely by private developers. Existing oversight mechanisms such as Meta’s user plebiscites and the Oversight Board have had limited impact due to low participation and institutional constraints. Meanwhile, policymakers have struggled to keep pace with rapid AI advances, with no significant federal legislation directly governing AI model constitution or alignment. This has contributed to public skepticism and a social license crisis surrounding AI development and deployment.
Analysis
Advocates highlight that traditional legislative bodies are institutionally ill-equipped to handle complex, rapidly changing AI governance issues. The Council’s design leverages deliberative democracy principles, randomly engaging a representative public to balance inclusivity and expertise through structured education. The involvement of CAISI to independently evaluate constitutions and AI models offers a safeguard against biased or inadequate standards. The proposal posits that social license for AI development demands direct public input in foundational governance rather than reliance on private developers or elected officials alone.
Who Is Affected
- Federal government agencies procuring AI systems
- Developers of frontier AI models seeking federal contracts
- General U.S. public, whose values would influence AI governance standards
- AI research and nonprofit organizations submitting constitutions for consideration
What Remains Unclear
- Whether this proposal will be formally introduced or adopted by federal lawmakers
- The specific legal mechanisms required to condition federal procurement on constitution adoption
- How the Council will balance divergent public values in practice when selecting constitutions
- Potential challenges in coordinating with private AI developers resistant to mandated standards
- How CAISI will operationalize ongoing evaluation and amendment recommendations
What Comes Next
This information was not confirmed in the reviewed sources.
Sources
This article is based on reporting and publicly available information from the following source:
Read more AI Regulation stories on Goka World News.