News

The AI Governance Debate Moves Beyond Washington

• By John K. Waters
• 05/26/2026

The debate over artificial intelligence regulation is moving beyond technology companies and national governments, as religious leaders, military analysts, and civil society groups press for clearer limits on the use is AI in work, war, and public life.

Pope Leo XIV entered that debate this week with his first encyclical, a major teaching document that called for robust international regulation of AI. The document, titled Magnifica Humanitas, warned against the concentration of AI power in the hands of a small number of technology companies and criticized the use of AI systems in life-or-death decisions.

The encyclical adds a new voice to an AI governance debate that has often been centered on Washington, Brussels, Beijing, and Silicon Valley. Its focus is broader than product safety or data privacy. It frames AI as a question of human dignity, labor, war, inequality, and political power.

Pope Leo XIV warned against allowing AI systems to make life-or-death decisions and criticized what he described as a “culture of power” driving an AI arms race. The document called on developers and governments to prioritize ethical and spiritual values over profit and dominance.

The timing is notable. AI systems are being integrated into workplaces, software development, search engines, education, health care, and defense systems. At the same time, governments are struggling to define rules for model testing, liability, copyright, privacy, national security, and autonomous weapons.

Military use of AI has become one of the most contentious areas. AI is already being used in military settings, of course, including surveillance, object detection, and targeting-support systems. The modern U.S. military AI push found its way to Project Maven, a U.S. Department of Defense initiative launched in 2017 to use AI to analyze drone surveillance footage.

Project Maven evolved into the Maven Smart System, and companies including Amazon, Microsoft, and Palantir became involved after Google withdrew from the original effort following employee protests.

The issue has become more complicated as AI companies try to set their own use policies while competing for government contracts. Anthropic, for example, has sought to maintain restrictions on autonomous targeting and domestic surveillance, even as commercial and government pressure around military AI increases.

That tension points to a central governance problem: AI companies can write policies, but governments, militaries, customers, and competitors can create incentives to narrow or reinterpret them. The result is a fragmented system in which ethical limits may differ across companies, jurisdictions, and use cases.

Pope Leo’s encyclical speaks directly to that gap. The document criticized the concentration of digital power and urged stronger international regulation to prevent AI from worsening inequality, displacing workers without safeguards, and expanding the use of automated systems in war.

The document also situates AI governance within a longer historical context. It draws parallels with Pope Leo XIII’s Rerum Novarum, the 1891 document that addressed labor and social conditions during the Industrial Revolution.

That framing may prove influential because the AI debate is no longer only about how to regulate software. It's increasingly about whether existing institutions can manage a technology that could affect employment, information integrity, defense policy, education, and social trust simultaneously.

The policy response remains uneven. Some governments are pursuing safety testing and model oversight, while defense agencies are accelerating AI adoption. Companies are publishing safety frameworks, but critics say voluntary commitments are not enough as competitive and geopolitical pressures rise.

For the AI industry, the expanding governance debate creates a more complex landscape. Technical performance is still central, but companies are now being judged on transparency, labor impact, defense relationships, safety controls, and the social consequences of deployment.

For policymakers, the challenge is to decide whether AI can be governed through existing rules, voluntary corporate commitments, and sector-specific oversight, or whether the technology requires broader international standards. Pope Leo’s intervention suggests that pressure for the latter is likely to keep growing.