DC lobbyist Zak Kukoff on why data centers are becoming a top political issue ahead of 2028
Apr 7, 2026 with Zak Kukoff
Key Points
- Data center permitting has become a top 2028 campaign issue as bipartisan consensus forms around concerns over energy costs, water use, and local heat, with AI super PACs spending heavily on congressional races while avoiding pro-data center messaging.
- Democrats flipped Virginia's governorship and Georgia utility seats by campaigning on ratepayer protection, exploiting voter concerns that data centers drive energy costs up while creating few local jobs.
- Federal permitting consolidation for nuclear and clean energy is the critical unlock, as fragmented local approval authority can kill projects with national significance before the industry's window to build closes.
Summary
Data centers as a 2028 political liability
Zak Kukoff runs the tech and venture practice at Lewis-Burke Associates, a federal lobbying firm working across science, tech, healthcare, and education. His read on Washington right now: the data center buildout has quietly become the most charged political issue facing the tech industry, and the window to get infrastructure built is narrowing.
The political anatomy of the problem
A broad bipartisan consensus is forming around the idea that data centers raise energy costs, consume water, and increase local heat. Kukoff acknowledges much of this is perception rather than fact, but says that distinction is losing ground fast. Populist and opportunist candidates alike are recognizing data centers as a viable 2028 campaign issue, and super PAC money is already flowing in response.
Leading the Future, the OpenAI-affiliated super PAC, and Public First (the Anthropic-aligned equivalent) are both spending heavily on congressional races. Kukoff points to New York's 12th district as one example. In two recent Illinois races, Leading the Future won one of two contests, running ads about democracy and local candidates, not data centers. That's deliberate. Running pro-data center ads is not a winning strategy, so the PACs are trying to change the subject entirely.
The More Perfect Union account on X is, in Kukoff's view, the most effective political messenger operating today. Its formula is simple: local communities, portrayed sympathetically, pushing back against faceless tech firms. That framing is gaining traction faster than the industry's counter-messaging.
The ratepayer problem
Virginia is the sharpest example of where this leads. Kukoff says the data center question was a meaningful factor in Democrat Abigail Spanberger winning the recent Virginia gubernatorial race, after she campaigned on a straightforward kitchen-table message: energy costs are up, and she would slow data center permitting to bring them down. In Georgia, Democrats flipped Public Utility Commissioner seats on the same argument, and Kukoff thinks they now have a credible shot at the governorship for the first time in over twenty years.
The ratepayer protection pledge is one attempt to get ahead of this, but Kukoff is skeptical it gets codified into law. The more structural problem is messaging. Every time a prominent AI voice predicts mass job displacement or societal disruption, it undercuts any argument that local communities should welcome data center construction. His comparison: the architects of NAFTA did not campaign by promising offshoring. Tech, he argues, needs to stop doing that work for its opponents.
Tax revenue is one obvious local benefit, but it isn't landing. The counterargument, that data centers employ very few people and can be built somewhere else, is simple enough to stick.
Energy and the permitting bottleneck
On nuclear and clean energy, Kukoff sees federal permitting consolidation as the critical unlock. The problem is not energy production so much as the fragmentation of approval authority, local permitting bodies can effectively kill projects that have national-scale implications. A centralized, one-stop-shop permitting regime would reduce that exposure. He also flags co-location with stranded energy sources, places where generation already exists but transmission constraints prevent it from reaching demand, as a nearer-term opportunity, drawing a parallel to how crypto mining occupied similar sites for years.
Space-based data centers are not a credible near-term release valve, in his view. What's more likely to shift the political calculus is a crisis, a Sputnik-style moment that makes domestic AI infrastructure feel urgent rather than intrusive. He floats DeepSeek as a possible early version of that catalyst. The Germany-nuclear parallel is pointed: German politicians celebrated shutting down reactors until energy security became an acute problem overnight.
The core tension Kukoff leaves on the table is one of timing. The tech industry needs a window long enough to get data centers built and AI operationalized before adversaries pull ahead. Moving too slowly cedes that ground. Moving too aggressively alienates the communities that ultimately decide elections.