FERC pressures PJM to curb data-center cost spillover—while Congress fights over AI power bills

On September 30, 2026, the U.S. Federal Energy Regulatory Commission (FERC) asked grid operator PJM to revise a plan aimed at shielding households from the cost impacts of data center growth. In parallel, the U.S. Senate rejected a bill that would have targeted AI data center electricity costs, with Democrats arguing the proposal lacked enforcement “teeth” and calling for mandatory measures to rein in soaring consumer bills. The political fight is now spilling into the information domain: House Speaker Mike Johnson dismissed rising public opposition to data centers—citing concerns about water pollution and higher electricity bills—as a “Chinese psyop,” framing the backlash as foreign-influenced rather than locally driven. Together, the developments show regulators and lawmakers wrestling with how to allocate grid and environmental externalities from AI-driven load growth without triggering a consumer backlash. Strategically, the dispute is about domestic energy governance under stress from AI expansion, but it also carries geopolitical signaling. By attributting public resistance to a “Chinese psyop,” Johnson is effectively linking U.S. infrastructure politics to the broader U.S.-China technology and influence contest, potentially hardening positions on both sides of the debate. The beneficiaries are likely to be data center operators and grid planners seeking predictable cost recovery, while households and consumer advocates push for stronger protections and clearer accountability for environmental and utility impacts. FERC’s intervention suggests the regulator is willing to constrain how costs are socialized across ratepayers, which could shift bargaining power between utilities, PJM, and large load customers. If Congress fails to pass enforceable measures, the regulatory pathway may become the primary mechanism to manage AI load externalities. Market and economic implications are immediate for U.S. power pricing expectations, utility rate design, and grid investment incentives. If PJM’s plan is revised to better shield homes, it could reduce near-term volatility in retail electricity bills but may increase pressure on wholesale market participants or large-load customers to bear more of the costs. The legislative rejection also implies continued uncertainty for the policy framework governing AI data center energy consumption, which can affect demand forecasts used by power generators, transmission developers, and renewable project developers. In the short term, the narrative around “soaring electricity bills” can influence investor sentiment toward regulated utilities, grid equipment suppliers, and demand-response providers, while also affecting risk premia for companies exposed to ratepayer-funded infrastructure. While the articles do not name specific tickers, the most sensitive instruments would be U.S. regulated utility equities and power-related ETFs, where policy-driven expectations can move valuations quickly. Next, watch for PJM’s revised filing and FERC’s follow-up decisions, including whether the regulator requires a more explicit cost-allocation mechanism for data center-driven grid upgrades. In Congress, the key trigger is whether a new bill emerges with enforceable requirements—such as mandatory mitigation, caps, or standardized cost-sharing—rather than voluntary or weakly enforced provisions. The “Chinese psyop” framing raises the stakes for escalation in the political narrative; a shift toward bipartisan, evidence-based environmental and consumer protections would de-escalate the rhetoric, while further foreign-influence claims could polarize the policy process. Over the coming weeks, indicators to monitor include PJM’s load growth assumptions for AI centers, utility rate case filings referencing data-center impacts, and public testimony on water pollution and local environmental compliance. The escalation/de-escalation timeline will likely track the next FERC procedural milestones and the next legislative attempt to regulate AI power costs.

Geopolitical Implications

• 01

  Domestic AI infrastructure policy is being securitized through a U.S.-China influence lens.

• 02

  Regulatory constraints on cost recovery could reshape investment incentives for AI power demand and grid upgrades.

• 03

  Environmental externalities (water pollution) may become a durable political fault line, slowing permitting and compliance.

Key Signals

• —PJM’s revised filing and FERC’s response timeline
• —Whether Congress returns with an enforceable bill on AI data center electricity costs
• —Utility rate case language referencing data-center-driven grid costs
• —Public/environmental evidence on water pollution near data center sites