The Ai Power Race Favors U.s. States That Already Have Nuclear Plants

Google's recent agreement with Constellation Energy illustrates the importance of existing nuclear infrastructure. Under a 20-year arrangement, Google will purchase additional electricity generated through upgrades at 11 nuclear reactors across Illinois, Pennsylvania and New Jersey. The improvements are expected to add approximately 890 megawatts of capacity, with Constellation planning to invest more than $4.3 billion in the projects.

The arrangement demonstrates how technology companies can help finance electricity generation without waiting for entirely new power stations to be constructed. Upgrading an existing nuclear reactor, a process known as an uprate, can increase its output through improvements to equipment such as turbines, steam generators and control systems. Because the facilities already have operating licences, trained staff and grid connections, the process can be less complex than developing a new nuclear plant.

Google expects the additional capacity to become available in stages, with the full 890 megawatts projected to be online before the end of 2032. The company has also pursued other agreements intended to secure reliable electricity for its expanding AI infrastructure. These arrangements reflect the growing importance of long-term energy contracts in supporting data center development.

The geographic advantage is significant. According to the analysis published by Forbes on October 8, 2026, 28 US states have operating nuclear reactors that may offer opportunities for capacity upgrades, while 22 states do not have reactors available for this purpose. States with established nuclear infrastructure can potentially expand existing generation more quickly than regions that must develop entirely new sources of firm electricity.

Firm power is electricity that can be supplied consistently regardless of weather conditions. Nuclear plants can provide this reliability while producing electricity without direct carbon emissions during operation. Wind and solar remain important sources of clean energy, but their output varies with weather and time of day. A dependable electricity system therefore requires a combination of generation, storage, transmission and other resources capable of meeting demand when renewable production falls.

Colorado illustrates the challenges facing states without operating nuclear reactors. The state has expanded wind and solar generation, yet equipment failures at coal-fired power stations have exposed the difficulty of replacing large amounts of dependable electricity quickly. When a major generating unit becomes unavailable, utilities may have to extend the operation of older plants, purchase electricity from neighbouring systems or find alternative supplies.

New nuclear technologies could eventually offer another option, but small modular reactors and advanced reactor designs remain at different stages of development. Their ability to provide substantial commercial electricity will depend on construction costs, regulatory approval, manufacturing capacity and successful deployment. For states facing immediate power shortages, these technologies may arrive too late to solve near-term problems.

The AI industry's demand for electricity is also creating questions about costs for ordinary consumers. Long-term contracts can provide utilities with predictable revenue and help finance new capacity, but expanding data centers may increase competition for available electricity and grid infrastructure. Regulators must determine how costs are allocated so that households and existing businesses do not unfairly subsidise facilities built primarily to serve large technology companies.

Google has pursued a broader portfolio that includes existing nuclear plants, advanced nuclear projects and geothermal energy. This approach reflects the reality that no single technology will meet every region's needs. Areas with nuclear infrastructure may benefit from upgrades, while other locations could explore geothermal generation, renewable energy paired with storage, or carefully planned new power stations.

For policymakers, the challenge is to make reliable, low-carbon electricity available before demand outpaces supply. This requires predictable permitting, investment in transmission networks, workforce development and financing arrangements that encourage new generation without transferring unreasonable costs to consumers.

The expanding AI economy is making energy infrastructure a central factor in regional competitiveness. States that can deliver dependable electricity may attract data centers, manufacturing projects and investment, while those facing prolonged supply constraints risk losing opportunities. Existing nuclear plants provide an immediate advantage, but the longer-term challenge is to ensure that every region can expand its energy capacity sustainably as artificial intelligence continues to reshape the economy.