Clean Energy Investment Doubles but Grid Gaps and AI Demand Extend the Transition Timeline

Clean Energy Investment Doubles but Grid Gaps and AI Demand Extend the Transition Timeline Record renewable investment and IEA forecasts point to a faster energy shift, yet grid bottlenecks and AI power demand are complicating the arithmetic. Renewable energy investment is on track to hit $2.2 trillion this year, according to the International Energy Agency — more than double the capital committed to fossil fuels and representing over 40% of the $3.3 trillion estimated for the global energy sector.⁴ That number tells part of the story. The infrastructure needed to absorb the new supply is not keeping up. As Energy Monitor noted in its 2025 review, grids continued to lag even as clean energy accounted for the bulk of new capacity additions — a constraint "not rendering the energy transition moot," it observed, "but certainly raising concerns about the pace of change."⁵ Coal has not surged the way some predicted after the Iran war. Ember's analysis, shared exclusively with Carbon Brief, estimated the global rise in coal power output this year at no more than 1.8% in a worst-case scenario — well short of the more alarming projections that circulated when regional energy security became a political priority.² The IEA's Electricity 2026 report frames the longer arc. Global power demand is set to expand at more than 3% per year on average through the rest of this decade, driven by electrification, industrial load and the rapid build-out of data infrastructure. Coal's share of the generation mix is projected to decline, but the beneficiaries include nuclear and natural gas alongside renewables.¹ IEA forecasts show renewable output growing by roughly 1,000 TWh annually through 2030, with solar PV accounting for over 600 TWh of that gain. Renewables and nuclear combined are on course to reach 50% of the global power mix by the end of the decade.¹ The demand side is moving faster than the models expected. Artificial intelligence is the most discussed complication: BloombergNEF found in a recent report that data-centre energy use driven by AI will be "a key new source of electricity demand" through the coming decade, keeping fossil fuels in use for longer than the supply picture alone would imply. The IEA estimates AI and data centres could account for as much as 4% of global electricity consumption by 2030.⁴ Gas sits at the centre of that dynamic. As grids struggle to integrate intermittent renewable capacity, gas-fired generation is increasingly valued for its flexibility — able to start and stop quickly where solar and wind cannot. Energy Monitor described gas as offering "better grid stability than other energy sources." Its role is expanding even in scenarios where renewables are growing fastest.⁵ The developing world's trajectory adds further complexity. The Economist, in its survey of long-run demand scenarios, noted that the developing world's share of global energy consumption rises from 55% in 2021 to around 70% by 2050 across all three principal scenarios — and that the policy levers shaping that arc vary greatly from economy to economy.³ The track record on forecasting the energy mix is not reassuring. At an IEA meeting in 1983, assembled experts' predictions for oil demand in the following decade collectively missed the actual outcome. What the data show now is directionally consistent — the mix is shifting toward cleaner sources — but the pace of that shift, the enduring role of gas as a bridging fuel and the unplanned demand additions from AI infrastructure all suggest the timeline is longer than most scenarios have assumed.³ The next test is whether grid investment in the second half of this decade catches up with capacity additions. If it does not, low-cost renewable electricity will remain stranded at the point of generation, and gas will continue to clear the margin for years beyond what the investment headlines suggest.⁵