The Vatican stepped into the AI debate with a bold move: inviting an atheist AI executive to help unveil its first-ever encyclical on artificial intelligence. Christopher Olah, cofounder of Anthropic, sat beside Pope Leo XIV as the pontiff delivered a scathing critique of AI’s unchecked power.

Olah’s presence was no accident. The Vatican has spent years courting Silicon Valley’s AI minds, seeking a foothold in a field too complex for sermons alone. Anthropic’s emphasis on AI safety and ethical guardrails aligned with the Church’s call for outside oversight and moral accountability.

The encyclical, titled *Magnifica Humanitas*, warns that AI’s immense power risks concentrating wealth and control in the hands of a few. It calls for “disarming” AI, demanding robust legal frameworks and independent oversight to prevent technology from becoming a tool of exclusion and surveillance.

Olah echoed that caution from inside the industry. He admitted that AI labs, including Anthropic, operate under incentives that can clash with “doing the right thing.” His blunt assessment: AI could displace human labor on a massive scale. The industry cannot police itself.

The Vatican’s approach is pragmatic. It does not expect the encyclical to halt AI progress or reverse layoffs justified by efficiency claims. Instead, it aims to spark dialogue and shame those pushing AI without regard for broader consequences.

Unlikely Allies in an Ethical Crossfire

Olah’s role is paradoxical yet pivotal. Raised evangelical but now an outspoken atheist, he once publicly challenged religious dogma. Today, he helps craft Anthropic’s AI “constitution,” a set of ethical principles guiding the company’s Claude chatbot. He even engaged Catholic clergy in shaping that moral compass.

This collaboration signals a new kind of alliance. The Vatican has moved beyond abstract moralizing to active participation in AI governance. Anthropic, founded by former OpenAI researchers wary of competitive excess, presents a rare Silicon Valley company openly wrestling with AI’s risks.

The encyclical sharply rejects the notion that AI ethics can be left to the industry alone. It warns that self-regulation is inadequate when power concentrates behind closed doors. Leo XIV insists that governments and civil society must reclaim control.

He also challenges military AI use, calling the Church’s centuries-old “just war” doctrine outdated. AI weapons must be disarmed, he says, directly confronting the Pentagon’s growing AI ambitions — a stance that puts the Vatican at odds with the Trump administration, which has taken a hard line against Anthropic.

Olah’s presence at the Vatican pulpit during this historic moment underscores a fundamental tension. He represents a company facing a U.S. government ban on military applications of its technology. Meanwhile, the Pope demands a pause and a reckoning with AI’s social costs.

The Vatican’s gamble is clear: by engaging insiders like Olah, it hopes to inject ethical urgency into Silicon Valley’s race for ever more powerful AI. Whether that will temper corporate ambition or simply add another voice to the cacophony remains to be seen.