Coalition for Digital Environmental Sustainability (CODES)

The Coalition for Digital Environmental Sustainability (CODES) is a global cooperation initiative co-led by diverse entities like Future Earth, German Environment Agency, International Science Council, Kenyan Tech Envoy, UNDP, UNEP and ITU. It was born from a mandate under the UNSG's Digital Cooperation Roadmap and was pivotal for aligning digital innovations with environmental sustainability, e.g by including the environmental sustainability principle in the GDC or pushing for the latest UNEA-7 resolution on AI and the Environment. From the CODES perspective, key outcomes of a successful Global Dialogue should include a shared recognition that AI governance must account for environmental sustainability as a key category alongside social and economic concerns. Hence, AI governance frameworks should clearly address both (1) the growing environmental footprint of AI systems and infrastructure, and (2) AI's potential to contribute to climate action and wider progress on the Sustainable Development Goals. This is also a clear mandate from the Global Digital Compact as well as the modalities resolution, that establishes the Global Dialogue "with a view to enabling AI to contribute to the implementation of the SDGs" - hence clearly including the 2030 Agenda's environmental dimensions.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Transparency, accountability, and human oversight

From a CODES perspective, the very obvious gap is addressing environmental sustainability as a - or within one - thematic area. This informs our choice: - "AI capacity-building" is urgent so that all countries can shape and align digital and environmental governance, assess environmental trade-offs, and use AI for sustainable development. - Governing "Social, economic, ethical, cultural, linguistic and technical implications of AI" is centre-stage for aligning the two mega-transformations of our Century: the digital and sustainability transformation. "Environmental implications" should also be included explicitly, but naturally interact heavily with almost all other dimensions mentioned. - "Transparency, accountability and human oversight" are essential to ensure AI-related environmental impacts are visible and managed. This includes transparency about model development and deployment footprints (energy, water and materials), procurement choices, and operational practices in data centres and supply chains.

As stated above, not including any reference to "Environmental Sustainability" - of AI as well as through AI - is a heavy gap. Formally speaking, environmental goals are a key concern of the 2030 Agenda. And it is incorporated deeply in the Global Digital Compact (as the central reference for the establishment of the Dialogue). The modalities resolution established the Dialogue to enable AI to contribute to the implementation of ALL the SDGs. While it is indeed unfortunate that "environment" was not included in the a-g list of the modalities resolution explicitly, this should not be a reason to neglect the key overall mandate of the Dialogue - and the very real-world importance this environmental perspective already has and will have. Including the immense environmental footprint of the digital and AI infrastructure (from raw materials, water, energy, pollution, soil, waste, etc.) up to the many digital and AI innovations that need to be steered towards environmental goals. Additionally, we want to flag "Financing and Incentives". The Global Dialogue on AI Governance should address aligning finance, public procurement and innovation policy to find pathways to prioritise positive high-impact AI use cases for sustainable development and low-carbon digital infrastructure best practice.

The absence of common measurement and disclosure norms makes it difficult for policy makers to make informed decisions. Monitoring the environmental impact of the ICT sector is crucial for informed decision-making, especially as the sector's emissions and energy consumption continue to grow. Clearer governance can accelerate transparent reporting and the adoption of standard-based best practices.

The AI Dialogue can advance international cooperation by acting as a bridge between existing sustainability and digital processes. It can also promote shared principles that integrate safety, human rights and environmental sustainability, support convergence on the disclosure of environmental footprint information; catalyse capacity-building partnerships and enable peer learning on regulatory and policy approaches. The Dialogue can also help identify gaps where coordination is currently weak and elevate them to the attention of relevant bodies.

The "Coalition for Digital Environmental Sustainability (CODES)" was from its start in 2021 and is an early driver of building bridges between digital and sustainability communities and policy processes. It is orchestrator and part of a growing nexus of initiatives including "Green Digital Action" - Green Computing track, or the "Coalition for Sustainable AI". The Dialogue could acknowledge and leverage the work done by different processes and the involved stakeholders.

It is a great opportunity for a new format such as the Dialogue to really innovate the way we do stakeholder engagement. The formats should refrain from long lists of statements that do not interact with one another or do not circle around a specific common interest. Instead, the Dialogue should celebrate different innovative formats, including experimental ones. More generally, the question around what the Dialogue itself is and can be - and what not - should be addressed early and openly. Clarifying the specific interface it creates, including its specific function, is key for stakeholders to understand their role and options. One option could be to focus around three interface roles: 1. First, "Interface Science–Public", using the Dialogue as the key forum to present and discuss the recommendations of the Independent International Scientific Panel on AI. 2. Second, "Dialogue as Coherent Filter and Elevator", highlighting the procedural interaction and communicative rationality of the gathering itself, serving to bring relevant new topics to the surface and sorting things. 3. Third, "Interface Public–Decision-Makers", as a very relevant necessity of stakeholder engagement to allow for channels towards meaningful decision-making bodies, be it via summaries and recommendation products or via direct reporting to relevant bodies, including the General Assembly and national parliamentary bodies.

While there are still many underrepresented voices, we want to highlight as CODES the necessity to align sustainability and environmental communities including policy makers, researchers, activists, IPLCs in this field with the global dialogue on digital transformation and AI governance. There is much room for mutual learning, including meaningful lessons from decades of environmental and sustainability governance.