Are there ethical frameworks specifically for evaluating AI's environmental impact?

Emerging Frameworks, Not Yet a Single Standard

Ethical frameworks addressing AI’s environmental impact do exist, though the field is best described as actively developing rather than settled around one universally adopted standard. Several major international organizations have incorporated environmental and sustainability considerations into their broader AI ethics guidance, while researchers and technical bodies have separately proposed more specific approaches for measuring and evaluating AI’s environmental footprint. Together, these efforts represent meaningful progress, but they fall short of a single, comprehensive, universally enforced framework.

This pattern is fairly typical of how ethical guidance for emerging technologies tends to develop — general principles tend to emerge before detailed, universally adopted technical standards.

International Guidance That Includes Environmental Considerations

UNESCO’s Recommendation on the Ethics of Artificial Intelligence, adopted by its member states, is one of the more prominent international documents to explicitly address this area. It identifies environmental and ecosystem sustainability as one of its core values, calling on AI actors to favor data, energy, and resource-efficient AI methods and to assess environmental impact throughout an AI system’s life cycle. As a recommendation adopted by UNESCO’s member states rather than a binding treaty enforceable on private companies, its practical impact depends significantly on how individual countries choose to incorporate its principles into their own domestic policy and regulation.

The OECD has similarly incorporated sustainability-related considerations into its broader AI principles, which have been influential in shaping how many governments think about responsible AI development, including environmental dimensions alongside other values like fairness, accountability, and human rights.

More Specific Technical and Measurement Frameworks

Beyond these high-level international frameworks, researchers and technical organizations have proposed more specific approaches aimed at helping organizations actually measure and report AI’s environmental footprint — such as standardized ways of reporting energy consumption or estimated carbon emissions associated with training and running AI models. These efforts are valuable because high-level ethical principles alone don’t provide the technical specificity needed for consistent measurement and comparison across different organizations and systems. That said, adoption of any particular technical measurement framework remains inconsistent across the AI industry, and there isn’t yet a dominant, widely agreed-upon standard that most companies report against.

Bottom Line

Ethical frameworks addressing AI’s environmental impact do exist — most notably UNESCO’s global AI ethics recommendation and OECD guidance, both of which explicitly incorporate sustainability considerations — alongside various more technical measurement frameworks proposed by researchers, but no single, universally adopted and enforced standard currently governs how AI’s environmental impact should be assessed across the industry.

Go deeper

• Are There Industry Standards for Measuring AI’s Environmental Impact?
• What Does ‘Sustainable AI’ Actually Mean in Practice?
• What Efficiency Improvements Are Reducing AI’s Environmental Footprint?