AI has come, seen, and conquered the enterprise agenda. And in its next phase of growth, physical capacity for growth will be as crucial as model capability. US data center power demand is forecasted to more than double, from 31 gigawatts in 2025 to 66 gigawatts in 2027. This insatiable demand for energy but also water and land has created a rare universal agreement. Governments, regulators, utilities, communities, and companies all face a perfect storm that must answer: Who gets scarce power, who bears the responsibility for optimization of AI infrastructure, who absorbs risk, and who benefits? No organization can ignore costs, delays, outages, stranded assets, and community opposition that AI couples with all data center investments.

As the UN warns that global warming has exceeded 1.5 degrees Celsius, the sustainability conversations are shifting from avoidance to managing the overshoot. So what is in stock for 2027 at the intersection of AI, sustainability, energy, and regulations? Here is a sneak peek at three of our predictions for 2027:

Want grid access? It will cost you, as countries will start to levy tariffs on data center operators and builders.

With unprecedented growth demands, utilities cannot assume that every proposed data center will arrive on schedule or consume its promised load. Building generation and transmission for speculative demand could leave households and businesses paying for stranded infrastructure. Already, the White House has called on hyperscalers to procure their own power, fund upgrades, and pay for reserved capacity even when unused. In 2027, at least two countries will turn this ratepayer-protection principle into tariffs or equivalent obligations. Cloud and colocation costs will become more location-specific, contractually committed, and exposed to power-market risk.

Expect governments (and the utilities they own or regulate) to impose long-term minimum payments, financial guarantees, energy tariffs, and flexible-load requirements on data center operators. Cloud economics will consequently become more dependent on location, electricity contracts, grid maturity, and providers’ ability to generate or reduce power. Enterprises rapidly scaling AI capacity will increasingly inherit these costs.

Community-impact reviews will be the gate for new data centers.

The cloud and data centers are no longer abstract entities. Communities increasingly see, feel, and pay for the physical footprint behind the cloud: new substations, transmission lines, water withdrawals from farmlands, backup generators, persistent low-frequency noise, construction traffic, land conversion, air pollution from backup generators and dedicated gas-fired power, and higher utility bills. Cooling operations can strain local watersheds and aquifers, and without adequate pretreatment and discharge wastewater, that can burden municipal treatment systems. A Gallup poll found that 71% of Americans oppose an AI data center being built in their area, with concerns centering on resource consumption, pollution, public costs, and quality of life. And this resistance has consequences: Lawmakers introduced at least 375 data center bills in 2026, with roughly three-quarters seeking to restrict development or its costs. At least 75 projects worth approximately $130 billion have been disrupted by local opposition in Q1 2026.

Simply talking about social causes such as charities in communities that data center operators will champion will not be enough. Data center operators and builders will increasingly need to demonstrate both the technical and socioeconomic metrics of data center impact. Examples include who pays for grid upgrades, how water and emissions will be managed, what enforceable benefits communities receive, and how operators will mitigate noise, traffic, and emergency-response risks.

Digital product passports will survive regulatory retreat.

Companies waiting for Europe to abandon digital product passports (DPPs) are making a precarious bet. The European Commission already opened its registry and testing environment in July 2026, ahead of the February 18, 2027 deadline for certain large batteries. Requirements may simplify, but delay looks increasingly unlikely. Manufacturers must now connect fragmented product, supplier, material, repairability, and lifecycle data.

The strategic implication extends beyond compliance. DPPs will force manufacturers to connect product, supplier, material, repairability, and lifecycle data that often remains fragmented across enterprise systems. Waiting will turn a manageable data transformation into a last-minute compliance scramble.

Resource optimization becomes everyone’s business.

Remember the time when we started to switch off the water tap if it wasn’t in use, without wondering if it was an act of sustainability, cost optimization, resilience, or conservation? In 2027, unlimited AI ambition will face a reckoning with limited resources. Winners will not debate or care whether to label the response “sustainability.” They will be forced to plan AI, energy, water, land, infrastructure, and community impact together, because that is now what resilient growth requires.

Forrester clients can read our full Predictions 2027: Environmental Sustainability report to get more detail about each of these predictions, plus two more bonus predictions. Set up a Forrester guidance session to discuss these predictions with me and other key contributors of this report to plan out a 2027 environmental sustainability strategy and roadmap that will set your organization up for success.

Share