The Pope Spent Five Years on AI. Trump Couldn't Spare Ninety Days.

One week, two answers to the same question: who's looking out for ordinary people as the machines arrive.

On May 21, the White House had a signing ceremony scheduled. Tech executives gathered. An executive order was ready. It would have established a voluntary process allowing federal agencies up to 90 days to review advanced AI models before public release. Hours before the signing, it was pulled. President Trump explained his reasoning: “I really thought it could have been a blocker.” What he meant: any pause, any review, any look at what these systems can do before they are released, risks falling behind China. Whether that framing serves ordinary Americans or the companies racing to dominate a global market is a question Washington is not asking.

Four days later, Pope Leo XIV published Magnifica Humanitas, a 42,000-word encyclical on artificial intelligence and the human person. He had been working on it for five years.

Two events. One week. A clear picture of who is thinking carefully about what AI will do to ordinary people, and who isn’t.

What the Pope Said

Leo’s argument covers three distinct dangers.

On work: AI is already reversing the proper relationship between tools and people. Rather than machines being designed to serve workers, workers are being forced to adapt to the speed and demands of machines. Eliminating jobs in the name of profit is not progress. It is a choice that treats people as means rather than ends, and it will produce not just unemployment but what Leo calls “human and cultural impoverishment.”

On surveillance: AI gives whoever controls the data, government or private actors, an unprecedented capacity to monitor, track, and control populations. That capacity is not theoretical. ICE is already using AI tools to track people at scale, and the current administration is building the infrastructure to significantly expand that capability. The question the Pope is asking is who decided this was acceptable, and what legal framework governs it.

On warfare: machines that can decide to kill without meaningful human accountability are not simply better weapons. They are a different kind of decision, one that removes the human being entirely from the moral act. Leo calls for AI to be disarmed, drawing a direct parallel to the Church’s work on nuclear weapons. Anthropic, one of the leading AI companies, refused government demands to deploy its models for military and surveillance purposes. The government found willing vendors elsewhere without difficulty.

His demand across all three: voluntary corporate ethics are not enough. Robust legal frameworks, independent oversight, and governments that do not abdicate their responsibility are required. He published on the 135th anniversary of Rerum Novarum, the Church’s original response to the industrial revolution.

We Have Been Here Before

The industrial revolution transformed work at a scale no existing institution was prepared to manage. Workers who demanded an eight-hour workday in the 1870s did not get it until the 1930s. Children comprised 6% of the national labor force in 1900. When workers struck, the federal government dispatched troops to break them. The reforms that eventually protected ordinary people came only after sixty years, a depression, and sustained political pressure that those in power had spent decades resisting.

What Washington Is Doing Instead

On his first day in office, President Trump revoked the Biden executive order on AI safety. States moved to fill the gap. The administration then moved to preempt state laws it deemed inconsistent with a federal framework that does not yet exist as binding law. The voluntary review order pulled on May 21 was not a regulation. It was a request that agencies be allowed to look at powerful AI models before release.

AI executives have donated millions to President Trump, and industry-backed PACs have spent heavily to keep federal oversight off the table. The industry shaping this moment has plenty of influence in Washington. No institution accountable to ordinary people has been asked to do the same.

What This Means

The question is not whether AI is good or bad. The benefits are real. But the dangers the Pope identifies, to workers, to people subject to government surveillance, to the basic question of who decides when a machine can take a life, are also real, and they are arriving faster than any democratic response.

The Pope delivered a universal message: AI must be subordinate to human dignity. Let’s learn from history and not wait 60 years for Washington to listen and act.

Endnotes

1. Pope Leo XIV publishes Magnifica Humanitas, May 25, 2026: Vatican News, Pope Leo XIV’s Encyclical Magnifica Humanitas on AI

2. Trump pulls voluntary AI executive order, May 21, 2026: Axios, Anti-‘doomer’ feedback derails Trump’s AI executive order

3. Trump revokes Biden AI safety executive order, preempts state laws: TechPolicy Press, The Need for AI Regulation Is Greater Than Ever

4. ICE use of AI tools for tracking and surveillance at scale: American Immigration Council, Mission Creep: AI Surveillance at DHS Crosses Dangerous Line Into Tracking Americans

5. Anthropic refuses Pentagon demands; government finds willing vendors: CNN Business, Anthropic rejects latest Pentagon offer; Electronic Frontier Foundation, The Anthropic-DOD Conflict

6. AI executives’ political donations and industry PAC spending: TechCrunch, OpenAI’s vision for the AI economy

7. Industrial revolution labor history: U.S. Bureau of Labor Statistics, Labor Law Highlights 1915–2015

8. Rerum Novarum 135th anniversary context: America Magazine, In Magnifica Humanitas, Pope Leo delivers on a people-first vision for AI