What ethical frameworks address human dignity in AI design?

International Guidance Leading the Way

A number of prominent ethical frameworks address human dignity in AI design, with international guidance generally leading the way in explicitly naming and elaborating on this value. These frameworks vary in how directly they use the term “dignity” versus addressing closely related concepts, and they generally function as high-level guiding principles rather than detailed technical specifications for exactly how dignity should be operationalized in any specific AI system’s design.

Looking at a few of the most prominent frameworks helps illustrate both how seriously this value is treated at the international level and the practical limits of translating broad ethical principles into concrete design requirements.

UNESCO’s Explicit Treatment of Dignity

UNESCO’s Recommendation on the Ethics of Artificial Intelligence, adopted by its member states, stands out for explicitly naming human dignity as one of its core, foundational values underpinning the entire framework. The recommendation connects dignity to related principles including human rights, human oversight, and the idea that AI should not be used in ways that undermine human autonomy or reduce people to mere instruments of a system’s objectives. As with most such international recommendations, its practical force comes through member states’ adoption of related principles into their own domestic policy, rather than through direct, binding enforcement against individual companies.

Human Rights-Based Approaches

Beyond frameworks that name dignity explicitly, many AI ethics approaches are grounded more broadly in international human rights law, which itself is often understood as resting on an underlying commitment to human dignity as a foundational value from which specific rights are derived. Frameworks built around this human rights-based approach to AI ethics — addressing concerns like non-discrimination, privacy, and freedom from arbitrary interference — connect to dignity concerns even when dignity isn’t always the headline term being used, since violations of these more specific rights are frequently understood, at a deeper level, as violations of the underlying dignity those rights are meant to protect.

The OECD and Other Contributing Frameworks

The OECD’s influential AI principles, while perhaps less explicitly centered on the specific term “dignity” compared with UNESCO’s recommendation, incorporate closely related considerations through their emphasis on human-centered values, fairness, and ensuring AI systems work for the benefit of people and the planet. Various national governments and individual organizations have also developed their own AI ethics guidelines that incorporate dignity-related principles, though the specificity, rigor, and enforceability of these documents vary considerably.

Across all of these frameworks, a consistent pattern holds: dignity and closely related values are treated as important, foundational considerations, but the frameworks generally stop short of providing detailed, universally applicable technical specifications for exactly how every AI design decision should account for dignity, leaving considerable room for interpretation and, in practice, uneven implementation.

Bottom Line

Several prominent frameworks address human dignity in AI design, most explicitly UNESCO’s global Recommendation on the Ethics of Artificial Intelligence, alongside broader human rights-based approaches and the OECD’s AI principles, which address closely related considerations — together providing meaningful high-level guidance, though generally without the kind of detailed technical specification needed to fully standardize how dignity should be operationalized across every AI system’s design.

Go deeper

• What is an ai bill of rights and has any government actually adopted one?