Global Index on Responsible AI: 2026 Report | AI Research

The Global Index on Responsible AI: 2026 Report provides a comprehensive assessment of how nations are translating their commitments to responsible artificial intelligence into practical, enforceable reality. By grounding its analysis in human rights-based frameworks like the UNESCO Recommendation on the Ethics of AI, the report evaluates the current state of global AI governance, identifying a significant gap between the adoption of policy frameworks and the implementation of meaningful, protective mechanisms for citizens.

Assessing Global Governance

To measure progress, the research team utilized a global network of 135 country-level researchers to analyze 68,138 data points collected between November 2023 and September 2025. The index ranks countries on a scale of 100 based on 38 indicators across three core pillars: government policy and implementation, civil society engagement, and enabling conditions. These indicators are evaluated across five critical dimensions: Inclusion and Diversity, Ethics and Sustainability, Labour and Skills, Trust and Safety, and AI Use in Public Service.

The Gap Between Policy and Protection

The findings reveal that while AI governance is expanding—with 126 out of 135 countries now having at least one government policy or initiative—this growth does not necessarily equate to robust protection. A stark disparity exists between the Global North and the Global South: while the Global South has seen a surge in new framework adoption, 78% of these frameworks remain non-binding, compared to 42% in the Global North. This suggests that many countries are adopting policy language without establishing the legal requirements necessary to hold stakeholders accountable.

Transparency and Government Accountability

A major concern highlighted by the report is the lack of transparency regarding government-led AI projects. Although "Transparency and Explainability" is the most common policy area, with 58% of countries having some form of framework, only 18% of nations require the public disclosure of the algorithms used by their own governments. Furthermore, the report identified credible evidence that 35 countries are currently deploying AI systems categorized as "unacceptable-risk," indicating that government commitment to responsible AI often fails to apply to their own internal operations.

Moving Toward Enforceable Oversight

The report concludes that the current trajectory of AI governance is insufficient. To move beyond mere framework adoption, the authors argue that countries must prioritize three essential elements: enforceable, rights-based protections, the establishment of properly resourced oversight institutions, and the creation of accessible redress mechanisms for those harmed by AI systems. Without these concrete steps, the global expansion of AI policy risks remaining a symbolic gesture rather than a functional safeguard.