Major US technology companies have stayed away from a UN-backed AI climate pledge at COP31, raising concerns over accountability for rising data-centre emissions.

Google, Microsoft And Meta Stay Away As COP31 AI Climate Pledge Struggles For Big Tech Support

The420 WebDesk
6 Min Read


Follow us on Google News


Follow The420.in on Google News

A United Nations-backed initiative to reduce the environmental damage caused by artificial intelligence has failed to secure support from major American technology companies, raising questions about whether the industry is prepared to address the growing energy demands of AI.

As of November 17, Chinese technology giant Alibaba was the only major technology company to sign the Antalya Artificial Intelligence Pledge, launched during the COP31 climate summit in Turkey.

Google, Microsoft, Meta, Amazon and Nvidia were notably absent from the list of signatories. The initiative attracted 90 endorsements at its launch, but only nine governments had formally joined.

The development highlights a widening gap between technology companies’ public climate commitments and their willingness to participate in international efforts to measure and reduce AI’s environmental impact.

FCRF Launches Certified Chief Risk Officer (CRO) Program for India’s Risk Leaders

Why Major Technology Companies Have Not Signed

The pledge was introduced on November 16 by Turkey, the host of COP31, and the International Telecommunication Union, a United Nations agency responsible for digital technologies.

It encourages governments and businesses to use cleaner electricity, improve energy efficiency and disclose the environmental impact of AI infrastructure.

However, participation remains voluntary, meaning companies are not legally required to follow its commitments.

Reuters reported that Google, Microsoft, Meta and Nvidia did not respond to requests for comment about their absence.

Amazon declined to comment directly on the initiative. Instead, it highlighted its existing commitment to achieve net-zero carbon emissions by 2040 and reduce the carbon intensity of electricity consumption by 65% by 2030, compared with 2019.

The companies have not publicly established a common reason for staying away. Their absence, therefore, should not be interpreted as coordinated opposition to the agreement.

The International Telecommunication Union said discussions with major technology companies would continue, with the pledge remaining open for signatures until COP32 in 2027.

South Korea was among the governments supporting the initiative, while the United States, China and the European Union had not joined.

How AI Data Centres Are Increasing Electricity Demand

Artificial intelligence requires enormous computing power, particularly when companies develop and operate advanced AI models.

These operations run inside data centres, which house thousands of powerful computers working continuously to process information.

Unlike ordinary office computers, large AI systems can require substantial electricity for both computing equipment and cooling systems.

A hyperscale data centre is an exceptionally large computing facility designed to handle enormous digital workloads, including cloud services and AI applications.

As companies expand these facilities, electricity demand can rise sharply, increasing pressure on power grids and creating environmental concerns when additional energy comes from fossil fuels.

A Harvard-led study published on November 16 estimated that 403 major American data centres generated approximately 2.57 crore tonnes of carbon dioxide emissions over a 12-month period ending in April 2025.

The estimated emissions were comparable to Denmark’s annual carbon dioxide emissions.

The findings provided further evidence of the environmental pressures facing the technology industry as investment in AI infrastructure accelerates.

However, the study examined data-centre electricity use broadly, not emissions attributable exclusively to artificial intelligence.

Why Voluntary Climate Promises Face Scrutiny

The Antalya pledge asks participating organisations to improve transparency around electricity consumption, water use and emissions.

It also encourages more sustainable equipment design, responsible disposal of electronic waste and greater access to environmentally responsible AI technologies in developing countries.

These commitments could help establish common expectations for an industry expanding faster than many existing environmental reporting frameworks.

However, environmental campaigners have questioned whether voluntary participation can deliver meaningful change without binding obligations.

Julie Cerqueira, chief programme officer at the Natural Resources Defense Council, argued that voluntary standards alone would be insufficient.

The debate follows earlier disputes over the environmental consequences of expanding AI infrastructure.

In the United States, legal challenges involving gas-powered electricity generation for data centres have demonstrated how rapidly growing computing demand can collide with local environmental concerns.

Similar questions are emerging in India, where proposed data-centre developments have attracted scrutiny over electricity consumption, water availability and their impact on surrounding communities.

For Indian policymakers, the COP31 disagreement raises an important question: whether environmental safeguards should depend primarily on corporate promises or independently measurable standards.

The Antalya initiative includes plans to track progress annually through 2030, but its effectiveness will depend heavily on participation and transparent reporting.

For now, the absence of several companies building some of the world’s largest AI infrastructure networks remains its most significant challenge.

What this means for you:

As AI services become more common, consumers should examine technology companies’ published electricity, water and emissions disclosures rather than relying only on climate promises. Communities near proposed data centres can seek information about their expected resource requirements and environmental approvals.