When people outside the energy sector think about the grid, it’s usually only after something goes wrong, such as a blackout or a surprise bill spike. Lately, the culprit for the public thinking about energy has come from a new source of consternation: data centers.
As consumers raise the red flag on data centers, politicians across the spectrum are rushing to respond. The message is simple and politically powerful: these tech companies should pay for necessary grid upgrades and everyday ratepayers should be shielded from higher bills.
Those goals are admirable and easy to sell, but without being careful the public posturing is crossing into an old, risky territory: trying to regulate what electricity uses are valid and which ones should be restricted.
Already, we’re seeing legislative and regulatory moves aimed at directly capping data center energy use. On their face, such proposals look sensible: slow the pace of new demand until the grid and communities are ready. But past experience warns that deciding which end uses “deserve” power and which don’t is a fraught policy path.
A History Lesson
The first moment the United States seriously wrestled with the limits of energy supply was the 1973 oil crisis. Governments responded with direct, visible rationing policies to conserve energy: gasoline limits, lower speed limits, even daylight saving time alterations. Those measures were widely framed as patriotic sacrifices and, for the most part, were accepted because alternatives were scarce.
Later in the decade, however, the debate shifted from practical rationing to moral exhortation. President Jimmy Carter famously asked Americans to change daily habits — the “wear a sweater and turn down the heat” moment — and it landed badly. People felt their everyday lives were being micromanaged by the government. What began as a temporary collective sacrifice turned into a cultural backlash against conservation framed as involuntary sacrifice. The lesson from that era is straightforward: policies that intrude on personal choices and daily comforts risk political and cultural rejection.
Similar dynamics are at play today. Well-intentioned campaigns to phase out gas stoves or limit gas hookups in new homes collide with public resistance because they touch personal freedoms and conveniences. Efforts to outright ban gasoline vehicles — as opposed to letting consumers choose EVs voluntarily — have likewise met political headwinds.
How does this manifest at the industry level? Let’s use cryptocurrency mining as an example. At the height of crypto’s rise, amid concerns that mining operations risked overwhelming local grids, policymakers asked whether governments should decide which digital activities are worthy of electricity.
Some local and state actions did curb energy-intensive mining in certain places — and New York’s 2022 restrictions on some Proof-of-Work operations became a prominent regulatory signal. But broader, sweeping bans and punitive approaches largely failed to become the norm (see the failure of the federally proposed Digital Asset Mining Energy Tax). People rejected the idea of the government picking winners and losers in electricity consumption.
Flash forward to today’s moment: Data centers
All of this sets the backdrop for the current debate over data centers. The public reaction is visceral: when people see major tech companies installing the energy hogs in their communities, they often get resentful or suspicious.
Why is my power bill going from $200 to $300 right when these data centers go up? Why should the eco-conscious fret about turning off their lights when the tech billionaires are building unprecedented energy consumption? When industry leaders respond defensively — such as OpenAI CEO Sam Altman did when comparing energy efficiency of data centers and AI tools to the amount of food a human who did similar work requires — it only deepens the public’s ick factor.
From a strictly energy-policy perspective, though, reflexively stifling an emerging sector because of anxiety about its power use risks being short-sighted for a number of reasons:
Scientific and technological progress has often come from unexpected research paths; if new computing and AI work can drive breakthroughs that benefit society, denying it access to energy would be a blunt tool with heavy collateral damage.
Eschewing technological advancement, even if we’re not entirely sure where it’s headed, opens the door to that advancement simply happening elsewhere in the world and eventually needing to play catch up at higher costs (see for example, the current rush for U.S. energy storage manufacturing to catch up to China’s capabilities).
Do we risk getting stuck in looking at the grid in how it looks today vs. imagining how a future power system could look with major load centers being treated as an asset rather than a liability? The narrative for EVs, for example, has increasingly shifted from looking at them purely as a problem to solve and more into an asset grid planners could tap into. Could data centers follow the same pathway?
The pragmatic path — and the one that acknowledges both risk and opportunity — is to design smart guardrails. Companies behind data centers largely recognize this; many are willing to contribute to grid upgrades or make investments that help the system. If structured properly, those investments can modernize the grid in ways that benefit everyone: new generation where it’s needed, resilience in being first responders to public safety power shutoffs (rather than being an onus to which residents can’t safely or conveniently respond), and improved distribution capacity. That turns a perceived burden into an asset.
The unarguable facts are these: the grid needs investment, new generation is required, and tech companies are willing to invest. Are we prepared to say “no” to investment simply because of the destination of the electricity? Or can we craft policies that channel that investment into upgrades that serve utilities, customers, and future innovation?
The AI industry is making a lot of claims that cannot be substantiated while expecting consumers and taxpayers to pay for their adventures. Caution is clearly in order.