Phoenix Center for Advanced Legal & Economic Public Policy Studies

Public debate over artificial intelligence ranges between visions of mass labor displacement to the technology being overhyped—from substance to spectacle. Using Wave 119 of the Pew Research Center’s American Trends Panel, I document a large spread between how workers assess AI’s impact on others and on themselves: nearly three-in-four workers are concerned about workers generally but only one-in-three translate that worry into personal concern, less than half.

The age of Artificial Intelligence is upon us. What should the policy response be? The Biden Administration advocated for an “all of government” regulatory approach to AI in the hopes of somehow controlling the technology. In contrast, the Trump Administration has embraced AI and wants to encourage its natural growth with a de-regulatory approach. Regardless of which policy approach one may prefer, given the vast economic impact that AI will have on the U.S.

The Federal Communications Commission collects annual data on broadband prices through its Urban Rate Survey, covering a wide range of speed tiers and providers across urban markets. In the right analytical hands, the data offer an opportunity to track broadband prices over time. Yet, price is only one measure of market performance. Experience indicates that the quality of broadband service—typically measured by download speeds—consistently improves over time.

This Policy Bulletin evaluates claims that Chattanooga’s EPB municipal broadband network generated large employment gains in Hamilton County (TN). First, this Bulletin shows that the source paper invoked to justify the employment narrative explicitly finds no statistically significant employment effects in urban counties; Hamilton County is an urban county. Second, using modern causal inference methods, this Bulletin directly evaluates labor-market outcomes in Hamilton County. Across employment and unemployment outcomes, no economically or statistically significant effects are found.

This paper uses new data from the U.S. Census Bureau’s Household Pulse Survey to examine demographic determinants of artificial intelligence adoption and uses, fear of AI’s career consequences, and active resistance to AI use. Regression analysis on a sample of 7,209 adults finds that AI use rises strongly with education and income and declines with age. In contrast, among those who have adopted AI, the breadth of use is driven primarily by education and age, with income playing little role beyond the adoption decision.

Affordability has emerged as a defining political issue of the moment, and broadband internet service affordability is no exception—especially with the expiration of low-income broadband subsidy programs such as the Affordable Connectivity Program in 2024. Policymakers and advocates frequently cite current broadband prices as evidence that affordability challenges are worsening, justifying calls for expanded subsidies and regulatory intervention. But are broadband prices actually rising?

The Federal Communications Commission’s Section 706 Reports are a useful source for information on broadband deployment, even if largely used by Democratic administrations to justify regulation. Armed now with the new broadband fabric data, which attempt to measure broadband availability down to the location level, the FCC’s staff is faced with some new complexities. Summary statistics provided in the Report require linking the finely-tuned broadband data to population statistics.

Congress is currently considering legislation which would require operators of smartphone app stores (namely Apple for iOS and Google with Android) to verify the age of all users as a pre-condition of using any individual app. In other words, rather than have each alleged problematic app be responsible for verifying the age of their respective customers, this legislation would have the federal government shift the burden—and, by extension, the assumption of full liability for the actions of third parties over which they have no control—onto app stores.

The rapid emergence of artificial intelligence technologies has prompted a familiar regulatory response: in the absence of a cohesive federal framework, states are rushing in to fill the void with a patchwork of local initiatives. But while some may argue that this patchwork is a noble fulfilment of federalism, the economics nonetheless argue for these state laws to be preempted and replaced with a single federal framework. We first looked at this problem back in 2008 when we published a paper entitled Developing a National Wireless Regulatory Framework: A Law and Economics Approach.

Efforts to regulate Artificial Intelligence at the state level are proliferating like mushrooms after it rains. Unfortunately, given the lack of a clear statement by Congress that the federal government can preempt such state AI laws, current legal options to stop the proliferation of state AI regulation are shaky at best.