CFJ Op-Ed: Pennsylvania's GRID Standards and the Fatal Conceit of AI Industrial Policy
- Jeffrey Depp
- Jun 8
- 3 min read
Pennsylvania’s GRID Standards and the Fatal Conceit of AI Industrial Policy
Committee for Justice Senior Counsel for Law and Policy, Jeffrey Depp, has published a new article in the National Law Review examining Pennsylvania’s recently announced Governor’s Responsible Infrastructure Development (“GRID”) Standards and their implications for artificial intelligence, energy policy, and economic growth.
The article, “Pennsylvania’s GRID Standards and the Fatal Conceit of AI Industrial Policy,” builds on Depp’s earlier work examining the relationship between AI development, electricity demand, and energy infrastructure. While Pennsylvania deserves credit for recognizing the enormous economic potential of AI and data-center investment, the article argues that the Commonwealth’s new framework risks undermining those benefits through unnecessary government mandates and centralized planning.
The Market Is Already Responding
The article highlights two recent studies that challenge many of the assumptions underlying calls for increased government intervention.
A recent PwC analysis found that the U.S. data-center industry supported approximately 5.5 million jobs and contributed nearly $927 billion to the nation’s GDP in 2024. These findings underscore the growing importance of data centers as critical infrastructure supporting artificial intelligence, cloud computing, advanced manufacturing, and other sectors of the modern economy.
At the same time, an E3 study found that data centers do not necessarily impose costs on ordinary electricity customers. Examining multiple utility territories, the report concluded that data centers generally paid more than their incremental cost to serve and, in several cases, actually placed downward pressure on retail electricity rates. In other words, market mechanisms—including utility tariffs, pricing structures, and long-term contracting arrangements—are already helping manage many of the challenges associated with growing electricity demand.
According to the article, these findings support a broader conclusion: markets are adapting to the demands of AI infrastructure without the need for extensive government direction.
Pennsylvania’s Mistake
Pennsylvania’s GRID Standards seek to attract data-center investment while imposing a variety of conditions related to energy sourcing, labor practices, environmental objectives, and community-benefit requirements. Most notably, the standards require participating projects to satisfy escalating “clean firm energy” targets over time.
The article argues that these mandates reflect what economist F.A. Hayek described as the “fatal conceit”—the belief that policymakers possess sufficient knowledge to direct complex economic systems from the top down.
Artificial intelligence, energy markets, and infrastructure development are all evolving rapidly. Policymakers cannot know today what generation technologies, grid solutions, or business models will prove most effective a decade from now. By imposing predetermined requirements, governments risk preventing the very experimentation and entrepreneurial discovery that drive innovation.
The article also draws on Public Choice economics, which emphasizes that political actors respond to incentives just as market participants do. Viewed through that lens, many of the GRID Standards’ labor, compensation, and community-benefit provisions appear less like neutral economic policy and more like efforts to satisfy politically influential constituencies.
A Better Path Forward
Rather than attempting to dictate how AI infrastructure should develop, the article argues that policymakers should focus on removing barriers to investment and production.
This focus includes streamlining permitting, accelerating infrastructure development, expanding generation capacity, and allowing energy producers and technology firms to determine the most efficient solutions through market competition. The article contrasts Pennsylvania’s approach with more market-oriented efforts to expand energy supply and reduce regulatory obstacles to AI infrastructure development.
As states increasingly compete for data-center investment, the article concludes that the central challenge is not attracting AI infrastructure but creating the conditions necessary for it to flourish. Policymakers should be wary of replacing market-driven discovery with political allocation.
Read the full article, “Pennsylvania's GRID Standards and the Fatal Conceit of AI Industrial Policy,” in the National Law Review.
