On May 25, Pope Leo XIV issued his first papal encyclical letter, “Magnifica Humanitas,” inspired by the Catholic Church’s moral and social teachings about what it means to be human and a decade-long dialogue between the Vatican and Silicon Valley on the ethical and social challenges posed by AI technologies on society.

Specifically, Pope Leo warns that AI “threatens to normalize an anti-human vision,” a world that risks reducing humans “to mere cogs in a system driven toward ever greater efficiency.” Further, Leo warned that the concentration of immense digital power in the hands of a few private companies must be actively countered (a position mirrored by Microsoft CEO Satya Nadella, who in a June 21interview with The Wall Street Journal advocated for “companies (to) avoid dependency on a small group of frontier models”).

He argues for an urgent need to regulate AI technologies. “In the age of AI and robotics, it is no longer possible to rely solely on the ‘invisible hand’ of the market.” Moreover, “it is not enough to invoke ethics in the abstract; robust legal frameworks, independent oversight, and a political system that does not abdicate its responsibility are required,” he wrote.

In fact, Leo advocates entrusting the United Nations with overseeing AI technologies, as they “are essential instruments for promoting a civilization of love, for they can foster dialogue among nations and promote the peaceful resolution of conflicts.”

Unlike when a pope is considered infallible when he expounds on matters of Roman Catholic theology, an encyclical on AI technologies is a papal opinion on secular matters. While recognizing and attempting to protect the dignity of humanity, the pontiff focuses heavily on the most pessimistic AI technology potentialities rather than anticipated societal benefits, including faster and less expensive pharmaceutical development and medical relief.

His commentary on relying solely on the “invisible hand” of the free market is a mystery to an astute observer of Western governments (including the United States), as the business community’s lived experience is that of ever-increasing levels of public taxation and regulation — far from a libertarian’s utopia.

Pope Leo’s call for increasing government regulation of AI technologies is likely to result in greater concentration of government regulatory power (including by repressive regimes), benefitting incumbents at the cost of new entrepreneurial entrants.

Finally, having the United Nations as the overseer of AI technology development and implementation sounds like a recipe for commercial disaster, given their failed record at “foster(ing) constructive dialogue among nations and promoting the peaceful resolution of conflicts.”

Yet, there are inherent issues involving evolving technology that create real-world regulatory problems. The “Collingridge dilemma,” first coined by David Collingridge in his 1980 book “The Social Control of Technology,” refers to a methodological quandary in which efforts to influence or control further development of a technology — in this case, AI — face a double-blind problem.

First, information problems cannot be easily predicted until technology is extensively developed and widely commercialized. Second, there is a power problem in that control or change of technology is difficult when the technology has been entrenched in commercial systems.

Likewise, the “pacing problem” refers to the idea that technological innovation is outpacing the ability of laws and regulations to keep up. Thus, because of the Collingridge Dilemma (and the related pacing problem), AI technologies cannot be addressed through traditional, time-consuming legislative and administrative rule-making processes (except in specific instances), as the technologies are evolving too quickly to be addressed through a traditional “command-and-control” regulatory process.

However, Pope Leo is correct in emphasizing the insufficiency of AI ethics in the abstract. A complementary array of private and public governance mechanisms will need to be instituted, including the development of tech industry voluntary guidelines/standards (later referenced in administrative rules), in conjunction with a “lighter touch” federal regulatory framework (including a reasonable frontier model notification timeframe for federal government national security review).

Further, from a private governance perspective, active stakeholder involvement, and industry and company accountability benefit from the implementation of AI code of conduct operational transparency. Let vigorous private enforcement of civil liability laws, the specter of adverse industry/company reputational effects, and the threat of the passage of onerous legislative oversight work in tandem with necessary public regulation to best adapt to the “unknowns” of the pacing problem in emerging AI technologies.

Thomas A. Hemphill is the David M. French Distinguished Professor of Strategy, Innovation and Public Policy in the School of Management, University of Michigan-Flint. He wrote this for InsideSources.com.