Zendoric
← Back to the day · September 25, 2026

AI takes over the debate at New York Climate Week

🕒 Published on Zendoric: September 25, 2026 · 03:17

✨ AI-generated · how it's made

This week, as world leaders gathered in Manhattan for the UN General Assembly, coinciding with New York Climate Week, the topic dominating every conversation among investors, policymakers and activists has been none other than artificial intelligence.

This week, as world leaders gathered in Manhattan for the UN General Assembly, coinciding with New York Climate Week, the subject dominating every conversation among investors, policymakers and activists was none other than artificial intelligence. According to the MIT Technology Review journalist who covered the event, there is growing tension over AI's real impact on the climate and on climate technologies, and positions vary widely depending on who is speaking: some highlight AI's potential to accelerate scientific research, others point to how Big Tech funding and attention is boosting energy startups, and others focus on the emissions-intensive expansion of natural gas now underway to meet the sector's electricity demand.

UN Secretary-General António Guterres underscored this intersection of AI and climate in his speech on the assembly's opening day: he warned that the climate crisis fuels instability and displacement, and that artificial intelligence "could help solve all these challenges, or could make them worse." The point carries particular weight because this is the first year the world has had to genuinely accept that climate targets are slipping away. A recent report from the United Nations Environment Programme (UNEP) notes that the world has essentially already passed the point at which it would be possible to keep warming below 1.5°C above pre-industrial levels. Given that the target has in practice been missed, the report argues for not only rapid and drastic cuts to greenhouse gas emissions, but also for turning to carbon removal to absorb emissions already released into the atmosphere.

The big question, the one repeated in every panel and conversation, is whether AI can really help with any of this. What is indisputable is that this highly energy-intensive technology is putting the spotlight on the need to expand electricity supply and strengthen grid reliability. The result is more attention and more money for energy technologies, some of which happen to be low- or zero-emission. Companies in the nuclear, geothermal, wind and solar sectors have signed deals with the likes of Google and Meta, which are looking to power their new or expanding data centers. According to data from Currence, a financial tracking firm covering the sector, global venture capital investment in climate tech reached $26 billion in the first half of 2026, up 55% from the previous year, with products and services aimed at data centers taking an enormous slice of that pie.

Not every sector benefits equally, however. A Semafor article on that same report notes that venture capital investment in carbon management and low-emission fuels has plummeted this year. These are important solutions for tackling climate change, but ones that perhaps cannot be "sold" as easily to a data center. And so far, the data center buildout has come with a notable emissions cost: a few years ago, Microsoft, Google and Meta had ambitious greenhouse gas reduction targets, but all three companies have seen their emissions grow, largely because of the data centers needed to support AI.

Optimistic voices are not lacking. Evelyn Wang, MIT's vice president for energy and climate, said during a panel that AI could accelerate progress in areas such as the search for new catalysts. Wang also told the Associated Press that data centers will not keep worsening climate and water problems forever, putting the timeline at roughly a decade before these facilities stop adding planet-warming emissions. In the meantime, though, a great deal of natural gas capacity is being brought online to meet the immediate demand generated by new data centers, and once built, those power plants have a lifespan of several decades. That is part of the reason public pushback against AI is growing: more and more communities are experiencing pollution and noise near these data centers and the plants that supply them with electricity.

Overall, the dominant impression gathered this week is that much of the climate sector is, at the very least, skeptical of AI. The UN's climate chief, Simon Stiell, put it bluntly in a speech: he warned that AI leaders are "on thin ice" when it comes to their social license to operate, and that they are "sinking" in terms of public support. According to Stiell, the big tech moguls need to start demonstrating why AI's benefits outweigh its soaring costs, and to do so with the benefit of the many in mind, not just a small minority. The message emerging from this Climate Week is clear: AI's promise as a climate ally coexists, with mounting friction, alongside the evidence of its energy footprint and its direct impact on the communities near the infrastructure that sustains it.

🔗 Related on Zendoric

Sources & references