Solar energy has surpassed coal as a power source in the United States for the first time on record, generating 12.8% of the nation's electricity in May 2026 compared to coal's 12.2%. This milestone arrives amid surging electricity demand from artificial intelligence and data centers, revealing a critical tension between the technology industry's renewable energy goals and the immediate carbon costs of deploying AI infrastructure at scale.

Why Is Solar Overtaking Coal Right Now?

The shift reflects years of accelerating solar deployment. In May 2026, the US generated a record 45.5 terawatt-hours (TWh) of solar energy, up 17% from May 2025. Coal production, by contrast, has collapsed. In the past five years, coal's share of US electricity generation fell from 19.7% in May 2021 to just 12.2% last month, even as coal output rose slightly in May 2026 to 43.4 TWh.

Solar and battery storage combined accounted for 90% of all new power capacity added to the US grid in the first quarter of 2026, according to the Solar Energy Industries Association (SEIA) and analytics firm Wood Mackenzie. The US now has enough solar installed to power approximately 50 million households, with projections suggesting that capacity will double to power 100 million households by 2034.

"Overtaking coal for the first month on record shows just how far solar has come, from a niche contributor to the third-largest and fastest-growing source of power in the US electricity system," said Nicolas Fulghum, a senior data analyst at Ember, a think tank focused on the clean energy transition.

Nicolas Fulghum, Senior Data Analyst at Ember

What's Driving the Surge in Electricity Demand?

Tech companies are the primary driver of rising electricity consumption. According to the SEIA report, electricity demand continues to surge, especially from technology companies seeking to secure power sources to meet the growing demands of artificial intelligence and the data centers that run them. This creates an unusual dynamic: as AI becomes more prevalent, the infrastructure supporting it requires enormous amounts of electricity, forcing companies to compete for renewable energy sources.

However, a new study published in Communications Earth and Environment reveals a troubling timing problem. Rapid AI deployment increases near-term pressure on global carbon budgets because the digital infrastructure supporting AI produces emissions before many system-level benefits are realized. According to the research, across various deployment scenarios, the accelerated rollout of AI yields a median cumulative carbon debt of 2.85 gigatonnes of carbon dioxide before annual savings exceed annual infrastructure-related emissions in late 2031. This suggests that while AI may eventually help reduce emissions through optimization and climate modeling, the upfront carbon cost of building and powering AI systems could consume a significant portion of the remaining carbon budget needed to limit warming to 1.5 degrees Celsius.

How Are States and Regions Responding to Solar Growth?

Solar expansion is geographically concentrated in specific regions. States won by President Donald Trump in the 2024 presidential election accounted for 74% of all solar capacity installed in the first quarter of 2026, according to SEIA data. Leading states for new solar additions include Texas, Florida, Ohio, Indiana, Michigan, Arizona, and Mississippi.

This geographic distribution reflects both economic incentives and existing infrastructure. Solar deployment continues despite policy headwinds in Washington. In August 2025, the Environmental Protection Agency canceled Solar for All, a $7 billion Biden-era grant program designed to help lower-income households install rooftop solar and reduce energy bills. Meanwhile, the Trump administration has allocated more than $700 million in federal funds to upgrade coal power plants and support coal exports, using wartime authorities under the 1950 Defense Production Act.

Steps to Understanding AI's Energy Trade-offs

• Carbon Debt Timeline: AI infrastructure creates immediate emissions through manufacturing, transportation, and installation of data center equipment and power systems, but the emissions-reduction benefits from AI applications may not materialize for several years, creating a carbon deficit in the near term.
• Grid Decarbonization Speed: The faster the electrical grid transitions to renewable sources like solar and wind, the sooner AI's operational emissions decline; conversely, AI deployed in regions still dependent on fossil fuels carries a higher carbon footprint.
• Deployment Acceleration: Rapid scaling of AI systems increases the total embodied emissions from infrastructure before system-level savings can offset those costs, making deployment speed a critical variable in the carbon equation.

The solar milestone comes as China faces its own climate challenges. In early June 2026, China Southern Power Grid reported a record electricity load of 259 gigawatts driven by widespread cooling demand from high temperatures, breaking seasonal patterns where peaks typically occur in June and July. Simultaneously, China experienced heavy rains across multiple provinces, with nearly 10,000 residents evacuated in Guizhou after torrential flooding. Meteorologists attributed the unusually early and intense weather to shifting patterns reflecting broader climate change, with concerns that extreme weather could damage crops and rice fields across southern China.

The convergence of solar's historic milestone and AI's energy demands underscores a fundamental challenge for the technology industry: achieving climate goals requires both renewable energy deployment and careful management of the carbon costs embedded in the infrastructure that supports AI systems. While solar's growth demonstrates that clean energy can scale rapidly, the timing mismatch between AI's immediate emissions and its delayed climate benefits means that simply connecting data centers to renewable power is not sufficient to address the technology's full climate impact. The next five years will be critical in determining whether AI can deliver on its promise as a climate solution or whether its infrastructure demands will consume the remaining carbon budget needed to prevent catastrophic warming.