The AI & Democratic Values Index 2026 is the annual flagship report of the Center for AI and Digital Policy (CAIDP), an independent civil-society research organization headquartered in Washington, D.C. It is a cross-national tracker of national AI policies, regulatory frameworks, and adherence to democratic values across (typically) 80+ jurisdictions, and is the 2026 edition in the annual CAIDP series. The report is widely cited by intergovernmental institutions, civil-society coalitions, and national policymakers.
Publisher: Center for AI and Digital Policy (CAIDP) Released: 2026 ISBN: 979-8-9956958-0-6 Source: caidp.org
Scope and structure
Per the typical CAIDP structure, the 2026 edition covers:
- National AI policies and strategies by country
- Regulatory framework status — proposed, enacted, or enforced
- Adherence to democratic-values principles — transparency, accountability, oversight, rights protection
- Multilateral participation — engagement with UNESCO, OECD, G7, G20, GPAI, and the Council of Europe (COE) AI Convention
- Civil-society engagement in national AI policymaking
- Areas of concern — regulations or practices that score poorly against the democratic-values benchmark
The report provides a cross-national synthesis layer for AI policy: each country's posture, how jurisdictions compare, and where gaps and overlaps fall. This complements per-jurisdiction documentation of specific laws, executive orders, and frameworks (such as CA SB 53, the EU AI Act, EO 14365, and China's Algorithmic Recommendation Provisions) rather than duplicating it. Coverage extends beyond the US-China-EU triangle to South America, Africa, Southeast Asia, and the smaller European jurisdictions.
CAIDP is one of three civil-society organizations — alongside the Future of Life Institute and Stanford HAI — whose annual reports are cited across global AI-policy discourse. Where FLI focuses on safety and HAI on technical capability and adoption, CAIDP focuses on policy, democratic values, and human rights. The three reports are framed by CAIDP and others as complementary cross-national references.
Relation to other indexes and instruments
CAIDP ranks national governments on democratic-values adherence, a different unit of analysis from FLI AI Safety Index Winter 2025, which ranks frontier labs on safety. The Stanford HAI AI Index Report 2026 is more capability- and adoption-focused, while the CAIDP Index is more policy-focused; the two are described as triangulating.
The Index assesses national policy artifacts from a democratic-values lens, including the America's AI Action Plan. It also tracks multilateral instruments, among them the The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), Frontier AI Safety Commitments (Seoul, 2024), Paris AI Action Summit Declaration (2025), G7 Hiroshima Code of Conduct for Advanced AI (2023), EU General-Purpose AI Code of Practice (Final Version, 2025), and Council of Europe Framework Convention on AI and Human Rights, Democracy, and the Rule of Law. Jurisdiction-specific frameworks within its scope include the EU AI Act (Regulation 2024/1689), Colorado AI Act (SB 24-205) and SB 25B-004 (Date Amendment), California SB 53 — Transparency in Frontier AI Act, Texas Responsible AI Governance Act (TRAIGA / HB 149) — Source Summary, Utah Artificial Intelligence Policy Act (SB 149), Virginia HB 2094 (High-Risk AI Developer and Deployer Act, vetoed), Illinois SB 3444 — Artificial Intelligence Safety Act, and New York RAISE Act (S. 8828). National regulators and intergovernmental bodies it tracks include G7 (Group of Seven), EU AI Office, CNIL (Commission nationale de l'informatique et des libertés), UK Information Commissioner's Office (ICO), Irish Data Protection Commission (DPC), Infocomm Media Development Authority (IMDA, Singapore), Cyberspace Administration of China (CAC), and Taiwan Ministry of Digital Affairs (moda).
Provenance and confidence
Confidence is rated medium. The CAIDP Index is a civil-society report with explicit normative commitments to democratic values; its rankings represent CAIDP's assessment rather than independent objective measurement. Specific data points (the status of a regulation, the signature of a multilateral instrument) are factual and verifiable, whereas rankings and adherence scores are interpretive.
The full report is distributed as a 24MB PDF, too large to summarize in full here. Specific assessments are intended to be pulled per jurisdiction as needed.
Relationships
- supports: AI Governance (umbrella), America's AI Action Plan (cross-national assessment), EU AI Act (Regulation 2024/1689), The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), Frontier AI Safety Commitments (Seoul, 2024), Paris AI Action Summit Declaration (2025), G7 Hiroshima Code of Conduct for Advanced AI (2023), Council of Europe Framework Convention on AI and Human Rights, Democracy, and the Rule of Law
- related: FLI AI Safety Index Winter 2025 (complementary lab-side index), Stanford HAI AI Index Report 2026 (complementary capability-and-adoption index), CDEI and DSIT (UK), G7 (Group of Seven), International Association of Privacy Professionals (IAPP) (privacy-policy adjacent organization), EU AI Office
- instance-of: Cross-National AI Policy Tracking