Why AI Is Dominating the Climate Week Conversation

Artificial intelligence has become the unavoidable subject at this year’s Climate Week in New York—not because there is agreement about what it will do for the climate, but because its effects could run in sharply different directions.

AI might accelerate climate research, improve the development of climate technologies, and help address challenges associated with a warming planet. It could also increase electricity demand, draw investment and attention away from other climate priorities, and encourage more fossil-fuel infrastructure to support the computing industry.

That tension is now central to conversations among policymakers, investors, advocates, and journalists gathered in Manhattan alongside the UN General Assembly.

The climate debate has reached a more difficult stage

The timing gives the AI debate added urgency. A recent report from the UN Environment Programme said that the world has nearly passed the point where it could keep warming below 1.5°C above preindustrial levels.

That assessment changes the terms of the discussion. Cutting greenhouse-gas emissions remains the primary task, but the climate response will also need to address emissions already released into the atmosphere. The report therefore points to the need for both rapid emissions reductions and carbon removal.

Against that backdrop, AI is being evaluated not as a distant technological possibility, but as a tool that might influence an increasingly constrained climate response.

The case for AI in climate work

Supporters of AI’s climate potential point to its possible role in advancing research and developing climate technologies. In principle, that could help scientists and engineers investigate climate-related problems or identify useful approaches in areas such as energy and carbon removal.

AI is also drawing investment and attention toward energy startups. That may help some companies develop technologies intended to reduce emissions or manage energy systems more effectively.

Those benefits, however, are possibilities rather than guaranteed outcomes. AI does not automatically produce climate progress. Its value depends on the problems it is applied to, the energy used to operate it, and whether the resulting tools lead to measurable reductions in emissions or other climate benefits.

The cost of powering the AI boom

The most immediate concern is AI’s demand for electricity. As computing needs grow, so does pressure to expand power generation and supporting infrastructure.

The climate implications depend heavily on how that electricity is produced. If demand is met through additional natural-gas development, the expansion of AI could reinforce fossil-fuel use even while the technology is being promoted as part of a solution to climate problems.

The debate therefore cannot be settled by asking whether AI is “good” or “bad” for the climate. The more useful questions are what AI is being used for, what energy powers it, and what climate results can be demonstrated.

Climate technology is competing for attention

The influence of AI extends beyond its direct electricity consumption. Its prominence can affect which technologies receive capital, media coverage, and political attention.

For climate companies, that shift may be positive if it brings new money and technical expertise into energy innovation. It may be less helpful if AI becomes a dominant investment theme and draws resources away from technologies whose benefits are less fashionable but necessary for reducing emissions.

This is one reason the conversation at Climate Week is not only about electricity demand. It is also about priorities. Investors and policymakers must decide whether AI is strengthening the broader climate effort or narrowing it around a small set of highly visible technologies.

Carbon removal raises the stakes

The discussion is particularly consequential because carbon removal is becoming part of the climate response as the 1.5°C goal moves further out of reach.

Carbon removal is intended to take greenhouse gases that have already entered the atmosphere and remove them. AI’s potential role in that work remains uncertain.

That distinction matters. AI may assist research into carbon removal, but it cannot substitute for the underlying work of reducing emissions. Treating AI as a solution in itself would risk confusing a tool with an outcome.

A credible climate claim about AI would need to show more than improved computing performance or increased investment. It would need to connect the technology to demonstrable progress: lower emissions, more effective removal, or a concrete improvement in climate resilience.

A question of governance, not just innovation

UN Secretary-General António Guterres highlighted the uncertainty in remarks on the first day of the UN General Assembly. “The climate crisis fuels instability and displacement,” Guterres said. “Artificial intelligence could help solve all these challenges, or it could make them worse.”

That framing places responsibility on decisions made by governments, companies, investors, and the public. AI’s climate impact will be shaped by:

• The energy sources used to power computing
• The applications that receive funding and policy support
• The transparency of claims about climate benefits
• The treatment of emissions from expanding infrastructure
• The balance between near-term emissions cuts and longer-term technological bets

Climate Week’s unusually loud AI debate reflects the fact that these choices are being made while climate targets are slipping further out of reach. The technology may become part of the response, but its contribution will depend on whether its real-world benefits outweigh the emissions and attention it consumes.

For now, the most defensible position is neither uncritical enthusiasm nor blanket rejection. AI should be judged by measurable climate outcomes—and by the full cost of producing and deploying it.

Source

• “AI is dominating the conversation at Climate Week,” published September 24, 2026, Technology Review: https://www.technologyreview.com/2026/09/24/1145048/ai-climate-week/