OECD AI Principles

International principles for AI that is transparent, robust, accountable, and centered on people.

The OECD AI Principles set out high-level expectations for trustworthy AI. They emphasize inclusive benefit, human rights and democratic values, transparency, robustness, security, safety, and accountability. The principles are useful when technical performance is only one part of the decision and organizations also need to consider how a system affects people.

Principles do not provide a ready-made verification workflow. Their role is to make the larger obligations visible and help teams ask better questions about responsibility and disclosure. We reference them because information-integrity systems should preserve human judgment, explain their limits, and support correction rather than hide consequential choices inside automation.