The project
About SAIF
Sustainable AI for Flourishing (SAIF): Principles, Technology, Governanceis a multi-year transdisciplinary research project led by Associate Professor Henrik Skaug Sætra at the University of Oslo, funded through the TRUST program. It develops and tests a philosophically rigorous, democratically grounded framework that re-centers AI development around ecological integrity and human flourishing.
Current AI development is predominantly driven by optimization metrics that fail to account for long-term ecological sustainability and human flourishing. While "green" or "sustainable" AI often focuses narrowly on human self-interest and efficiency, a deeper transition is required to align AI systems with planetary boundaries, democratic legitimacy, and increased sensitivity to the more-than-human. There is a critical gap between abstract ethical principles and the "hard" instruments of governance — law, procurement, and code. SAIF exists to bridge that gap.
Three pillars
I
A Deep Ecology framework for AI
Starting from Arne Næss' deep ecology — a philosophy with flourishing at its center and a broad account of who and what has a right to flourish — we explore key tensions, including the understudied role of technology in deep ecology and the reconciliation of its radical de-centering of the human with practical AI ethics and safety, to develop a novel ecophilosophical framework for sustainable AI.
II
Anticipatory ethics and governance
Deep ecology is a long-range approach: it requires moving beyond presentism to the short-, medium-, and long-term implications of AI. Using a diverse set of futures studies methods, the project deals explicitly with potential futures of AI, sustainability, and flourishing — contributing to anticipatory ethics and anticipatory governance, and producing an Epistemic Futures Atlas.
III
Practical frameworks and guidelines
The synthesis becomes practical tools for developers, users, and regulators: research-based but aimed at non-academics who need concrete guidelines, KPIs, and indicators for evaluating the sustainability of AI — up-front and in continuous monitoring. The assessment toolon this site is the first step, building on theAI ESG Protocol.
Timeline and outputs
- Year 1 — Foundations and mappingSAIF Foundations Report, initial value and governance maps, partner commitments and pilot protocols.
- Year 2 — Anticipatory ethics and governanceBeta Futures Methods Handbook, first instruments and participatory designs, theory-of-change work.
- Year 3 — Implementation cycleEpistemic Futures Atlas, two co-creation labs, two organizational pilots with registered protocols, Governance Playbook v1.
- Year 4 — Consolidation and uptakeGovernance Playbook v2, policy and procurement templates, and the SAIF Embedding Kit.
The project plans pilots with public organizations in Norway, focused on practical tools for evaluating AI systems in the acquisition phase and for assessing an organization's internal use of AI. Across outputs, SAIF follows open scholarship practices where feasible.