Utilities Explore GENCO Structure to Insulate Consumers from AI Power Costs

The electric industry is testing a structural workaround for the surge in AI-driven power demand. Regulated utilities face a dilemma: adding dedicated generation capacity for data centers to their rate base would increase bills for all customers, while excluding it leaves them out of the market. An affiliated independent generating company, or GENCO, offers a potential solution by housing this expansion outside the regulated utility, allowing the business to capture AI revenue without directly burdening the general consumer base.

This approach draws on historical precedents where independent generators operated separately from utilities, ensuring that if the new entity failed, the regulated utility and its customers remained ring-fenced from the losses. The strategy relies on the GENCO financing its own debt separately from the utility. However, the structure is not without financial complexities, as the utility's credit rating often underpins the GENCO's borrowing costs, effectively leveraging consumer-supported credit for the affiliate's benefit.

Contractual Risks Remain in Affiliate Structures

The safety of the ring-fence depends heavily on the terms of power purchase agreements between the utility and the GENCO. If the utility commits to buying power for a term longer than the actual lifespan of an AI project, it may be left holding costly contracts. Furthermore, if the GENCO underperforms, regulators must determine who absorbs the shortfall. Critics argue that AI firms have significant resources to influence political and legal outcomes, potentially leaving regulators without the technical depth to protect consumers from cost-shifting.

Financial contagion is another concern. Distress at the GENCO could impact the holding company that controls both the affiliate and the utility. Since the holding company finances the utility's equity, any financial strain could raise the cost of capital for the regulated entity. Additionally, management under pressure might find creative ways to shift funds to the struggling subsidiary, undermining the intended financial isolation despite legal ring-fencing.

Regulatory Scrutiny Focuses on Equal Treatment

An alternative model involves building a GENCO entirely outside the regulated utility to sell power directly to AI centers. This avoids rate-base issues but introduces operational risks. Regulators are likely to scrutinize any preferential treatment given to the affiliate. For instance, if the utility prioritizes restoring service to its affiliated GENCO and AI customer during a storm, it could violate the principle of equal service for all public utility customers.

The core challenge is balancing the economic opportunity of AI infrastructure against the stability of consumer rates. As detailed in utility earnings reports, the industry is navigating these tensions by designing complex corporate structures. The success of these models hinges on strict regulatory oversight and clear contractual boundaries to prevent the subsidization of high-growth tech sectors by traditional residential consumers.