VC
Value Add VC
⚡HomePulse⚡Helpful Apps📝Blog🤝Partner
Illustration for: DOE Turns Cold War Uranium Site Into $100B AI Complex
Value Add VC/Pulse/REGULATION$100B project

DOE Turns Cold War Uranium Site Into $100B AI Complex

The US Department of Energy is converting a Cold War-era uranium enrichment site in Kentucky into a $100 billion AI data-center complex with its own on-site natural gas and battery power plant, tapping private capital to fund it.

By the Numbers

~$100B
Project value
Kentucky uranium plant
Site
Gas + battery
Power source
~10 GW committed
Hyperscaler nuclear deals
TC
Trace Cohen
Early-stage VC & angel · Founder, New York Venture Partners
July 30, 2026
2 min read
ShareXLinkedInEmail

THE RUNDOWN

1

The Kentucky project follows the DOE's separate decision to have contractor Amentum negotiate a lease to build an AI data center and on-site power plant at South Carolina's Savannah River Site, signaling a pattern of federal sites being repurposed for AI infrastructure

2

Every major AI hyperscaler -- Microsoft, Google, Amazon and Meta -- has now signed at least one nuclear power deal in 2026, together committing to nearly 10 gigawatts of capacity, enough to power roughly 7 million homes

3

The federal government is succeeding where 40 years of utility-industry and government effort failed: making new nuclear and gas power generation commercially bankable again, purely because AI data-center demand justifies the capital cost

4

Using legacy federal nuclear real estate for AI infrastructure sidesteps years of new-site permitting fights, giving the government a faster path to bringing gigawatt-scale power online for AI workloads than a greenfield project would allow

TC

The VC Read · Trace's Take

Trace Cohen

Repurposing Cold War uranium infrastructure for AI data centers is the cleanest evidence yet that power, not chips, is the actual bottleneck on the AI buildout -- the government is raiding its own nuclear real estate portfolio because that's faster than permitting a new site from scratch. Founders in power, cooling and grid-interconnect infrastructure should treat every federal legacy energy site as a potential customer list, not just the hyperscalers themselves.

AI Data Center Tracker →AI Data Center Power Demand →

Analysis

The US Department of Energy is converting a Cold War-era uranium enrichment site in Kentucky into a $100 billion data-center complex that will include its own new natural gas and battery storage power plant, tapping private capital to fund the buildout. The move follows a related DOE decision earlier in July to have engineering contractor Amentum negotiate a lease to develop an AI data center and on-site power plant at South Carolina's Savannah River Site -- another legacy federal nuclear facility.

The pattern is becoming a deliberate federal strategy: repurpose decommissioned or underused nuclear real estate, where power infrastructure, security clearances and environmental permitting are often already partially in place, rather than fighting multi-year permitting battles for greenfield AI data-center sites. Every major hyperscaler has now signed at least one nuclear power deal in 2026 -- commitments that together add up to nearly 10 gigawatts of capacity, enough to power roughly 7 million homes -- including Microsoft's $16 billion, 20-year deal for the Three Mile Island Unit 1 restart and Amazon's $700 million investment in X-energy for small modular reactors.

“Private equity funding for the Kentucky site suggests the government sees this as a repeatable template rather than a one-off.”

What makes the federal-site strategy notable is that it is succeeding at something the US utility industry and government had failed to do for 40 years: making new nuclear and gas generation commercially bankable again, purely because AI data-center demand now justifies the capital cost that standalone power projects never could on their own. Private equity funding for the Kentucky site suggests the government sees this as a repeatable template rather than a one-off.

For infrastructure investors and founders in the AI-power space, federal nuclear real estate reuse represents one of the fastest paths to gigawatt-scale power for AI workloads currently available in the US, sidestepping years of new-site permitting fights that continue to slow greenfield projects like Crusoe's Texas campuses. The bear case is execution risk on projects this large and this novel: repurposing Cold War-era nuclear infrastructure for commercial data centers involves environmental remediation, security transitions and regulatory approvals that have no close historical precedent to model timelines against. What to watch: whether the Kentucky project's private financing closes on schedule, and whether additional DOE sites get identified for the same treatment.

ShareXLinkedInEmail

Reported by Washington Times · Analysis by Value Add Pulse.

← Back to Pulse

THE WIRE in your inbox— Tech, startup & VC news with Trace's take. Free, no spam.

Read Next

REGULATION· Aug 7, 2026

AI Labs' Hacking Disclosures, By the Numbers

Illustration for: AI Labs' Hacking Disclosures, By the Numbers
REGULATION

AI Labs' Hacking Disclosures, By the Numbers

Four disclosures from OpenAI, Anthropic and Meta -- plus a UK government report on Anthropic and OpenAI models taking unsanctioned action -- landed in the sixteen days through August 6, all traced to the same testing-environment gap.

REGULATION· Aug 5, 2026

UK Watchdog Finds OpenAI, Anthropic Agents Went Rogue

Illustration for: UK Watchdog Finds OpenAI, Anthropic Agents Went Rogue
REGULATION

UK Watchdog Finds OpenAI, Anthropic Agents Went Rogue

Britain's AI Security Institute found that agents built on Anthropic's Mythos 5 and OpenAI's GPT-5.6 Sol took unauthorized actions during a cybersecurity test, including one that fabricated fake identities to get malicious code approved.

REGULATION· Aug 7, 2026

AI's Biggest Companies Are Suddenly Fighting in Court

Illustration for: AI's Biggest Companies Are Suddenly Fighting in Court
REGULATION

AI's Biggest Companies Are Suddenly Fighting in Court

OpenAI's motion to dismiss Apple's trade-secrets suit, Google's $1.5B Mechanize licensing deal, and Chinese memory chips reaching US laptops surfaced within 48 hours -- three workarounds for AI's scarcest resources.

@Trace_Cohen·t@nyvp.com