Meta has added 650 megawatts of solar capacity to its energy portfolio, pushing its total renewable holdings beyond 12 gigawatts and placing the company among the world's largest corporate buyers of clean energy.
The addition is a direct response to the electricity demands of AI development. Training and running large language models requires sustained compute power, which translates into significant data center energy consumption — a challenge that has become both a business and reputational priority for Meta as it scales products including its Llama model family and AI assistant integrations.
A Solar Addition With Unanswered Details
650 MW is roughly enough generating capacity to power several hundred thousand average U.S. homes under optimal conditions. Added to a portfolio already exceeding 12 GW, it illustrates how large the energy procurement operations of major technology companies have become.
Meta has not disclosed the location or developer behind the new solar project, nor has it provided a timeline for when the capacity will come online. The company has also not specified whether the agreement is a power purchase agreement — the most common corporate renewable procurement structure — or a direct ownership arrangement.
For a single corporation, a 12-gigawatt-plus renewable portfolio represents years of sustained deal-making and reflects how fundamentally large-scale tech has converged with energy infrastructure development.
AI's Emissions Problem Isn't Going Away
The announcement arrives as the technology industry grapples publicly with the tension between AI ambitions and environmental commitments. Google reported last year that its greenhouse gas emissions had risen 48 percent over five years, attributing a significant portion to AI-related data center energy use. Microsoft has similarly acknowledged its emissions moved in the wrong direction despite long-standing net-zero pledges, due largely to AI infrastructure buildout.
Meta has framed renewable procurement as its primary mechanism for keeping AI expansion compatible with its sustainability targets. The company has committed to net-zero emissions across its value chain by 2030 and states it matches 100 percent of its global electricity use with renewable energy purchases.
Critics of corporate renewable accounting note, however, that matching consumption with renewable energy certificates does not guarantee a company is drawing clean power at every hour of every day. The geographic and temporal mismatch between when and where renewables are generated versus when and where data centers consume electricity remains an industry-wide challenge.
How Meta's Strategy Compares to Rivals
Meta is not alone in scaling clean energy procurement to justify AI expansion. Amazon, operating the world's largest cloud infrastructure through AWS, has repeatedly claimed the title of largest corporate renewable energy buyer globally. Microsoft signed a landmark deal with Constellation Energy to help restart a unit of the Three Mile Island nuclear plant in Pennsylvania specifically to supply its data centers. Google pursues a strategy it calls "24/7 carbon-free energy", aiming to match clean power consumption on an hourly rather than annual basis.
For Meta, solar has been a consistent cornerstone of its renewable strategy, supplemented by wind power agreements across the United States and internationally. Solar and wind power purchase agreements remain the dominant procurement vehicle, though the company has also invested in longer-duration storage and grid infrastructure in select markets.
Whether Procurement Can Keep Pace With Demand
To put Meta's portfolio in perspective: the entire installed solar generating capacity of many mid-sized countries sits in the 12 GW range. Whether the portfolio is sufficient to cover Meta's actual and projected AI workload consumption is less clear.
The company has not published forward-looking electricity demand estimates tied specifically to AI, making it difficult for outside analysts to assess whether procurement is keeping pace with consumption growth.
What This Means
Meta's latest solar deal signals continued procurement intent, but the trajectory of AI energy demand across the industry means capacity secured today may already be insufficient for infrastructure needs by 2027.