The Vatican just entered the AI race: Pope Leo XIV released Magnifica Humanitas, his first encyclical, focused entirely on protecting the human person in the age of AI 😳

And sitting inside the Vatican launch was Christopher Olah, Anthropic co-founder and one of the most important people in AI interpretability.

This was not “Church vs. AI.”

It was the Vatican saying something Silicon Valley does not love hearing:

Technical alignment is not enough.

You can make the model safer.

You can make the benchmark scores higher.

You can make the guardrails cleaner.

But if AI concentrates power, replaces work without a safety net, or turns human dignity into a rounding error, the problem is no longer technical.

It is moral.

A few details make this historic:

→ The encyclical was signed on May 15, the 135th anniversary of Rerum Novarum, the Church’s famous response to the Industrial Revolution.

→ Pope Leo called for robust AI regulation, transparency, worker protections, and limits on machines making life-or-death decisions.

→ He warned against AI concentrating power “in the hands of a few.”

→ Olah’s presence gave Anthropic exactly the kind of external ethical anchor it has been trying to build its brand around.

This matters because Anthropic gets proximity to moral legitimacy, while the Vatican gets a seat at the AI governance table.

And the AI industry gets a very uncomfortable reminder:

The next phase of AI will not just be decided by benchmarks, GPUs, and model releases.

It will be decided by trust, labor, power, dignity, and who pays the price when “progress” moves faster than society can absorb.

Turns out, the next AI moat may not be intelligence.

It may be trust.