Page 108 - AI Standards for Global Impact: From Governance to Action

AI Standards for Global Impact: From Governance to Action

14�4 Standards, policies and regulations for sustainable AI and
environment

Regulatory frameworks and international standards can help guide the sustainable deployment
of AI in the energy and ICT sectors. It will also be important to consider data transparency,
institutional cooperation, and the alignment of national policies with global climate goals, as well
as clearly define the various stakeholders' roles in addressing the environmental impacts of AI.

Colombia’s data centre regulation project aims to improve the transparency and environmental
accountability of digital infrastructure as part of the country's effort to collect data on energy
consumption and GHG emissions. Recognizing the importance of political will private-sector
participation, a call was made for regional cooperation and harmonized standards, particularly
for reporting frameworks that could be adopted across Latin America.

Brazil offered a national perspective in the lead-up to COP30, which it will host. Brazil is
committed to environmental protection and climate diplomacy, given the centrality of sustainable
digital transformation and AI governance in shaping Brazil’s COP30 priorities. From Brazil's
perspective, there is a need to consider embedding AI energy regulation into the broader
UNFCCC agenda and multilateral institutions like ITU could help to scale support for capacity-
building in emerging economies.

The African Telecommunication Union (ATU) underlined the regulatory and infrastructural gaps
in Africa that challenge sustainable AI deployment and called for fit-for-purpose standards that
reflect the realities of developing and least developed countries, especially where digital access
is still limited. ATU noted that the ICT sector can play a dual role in accelerating access to energy
through smart grids and reducing its own footprint, but only if regulators are equipped with
the right tools and data.

Considering the work of France and the EU on digital environmental regulation, referencing the
European Data Act and energy performance monitoring of data centres, binding obligations
for AI developers and data centre operators, including disclosure of energy intensity, carbon
footprint, and water usage were deemed to be important. The discussion also noted the
importance of alignment between national regulators and global standard-setting bodies like
ITU, recalling the importance of data measurement at the international level.

Some key takeaways include:

a) Data access remains a bottleneck. Several panellists echoed the need for greater
transparency on energy consumption and GHG emissions, with this data being essential
for developing effective regulations and standards.
b) Considering the dual nature of AI as a tool for energy optimization as well as a source
of energy demand, regulators could consider systems-level regulatory approaches that
integrate AI policy with climate and digital infrastructure policy.
c) Developing countries need support, and capacity building for policymakers and regulators
will be important, particularly in Africa and Latin America.
d) Calls were made to accelerate the adoption of relevant standards from standards bodies
such as ITU.
e) The session framed COP30 as a strategic moment to embed digital sustainability and AI
governance into global climate action.
f) Opportunities for ITU-T Study Group 5

96