Silicon Valley Engages Catholic Church on AI Ethics

The Vatican's engagement with Silicon Valley traces to around 2016, when tech leaders including LinkedIn co-founder Reid Hoffman and then-Google CEO Eric Schmidt approached French priest Eric Salobir at a Bay Area conference, leading to the first "Minerva Dialogues," annual closed-door meetings named for the Rome church, Santa Maria sopra Minerva, where they were first held, according to The Atlantic and Religion News Service. Hoffman recalled the discomfort of walking past portraits of historical inquisitors: "It feels a little bit weird to be walking in voluntarily past these." The relationship escalated sharply around Anthropic specifically: an October 2025 introduction connected Anthropic co-founder Chris Olah, an atheist seeking Catholic perspectives on AI ethics, to Catholic University of America theologian Charles Camosy, who brought in Santa Clara ethicist Brian Green and priest Brendan McGuire for an exploratory meeting in December 2025. That grew into multi-day sessions, including an April 2026 gathering where roughly 15 Christian leaders met directly with the engineers building Claude to discuss the model's behavior, including what Green described as "bad" personas that can deny wrongdoing or blame users.

When Anthropic published Claude's Constitution in January 2026, Green, Bishop Paul Tighe, and McGuire were listed among its external commenters. The relationship's practical stakes became explicit in March 2026, when Camosy organized an amicus brief from Catholic moral theologians backing Anthropic in its dispute with the US Defense Department, which had labeled the company a "supply-chain risk" after it refused to let Claude be used for autonomous weapons systems or mass surveillance of Americans. The engagement culminated on May 25, 2026, when Olah appeared alongside Pope Leo XIV at the unveiling of the pope's AI-focused encyclical, a striking placement for a company co-founder at a papal document release.

This matters beyond optics: Anthropic's faith-leader engagement is functioning as an input to actual product and policy decisions, feeding into Claude's published Constitution and into the company's litigation posture on military AI use, not just its public messaging. Teams tracking how frontier labs build and defend safety and alignment commitments should treat this as a real governance channel, not a symbolic one, alongside more familiar mechanisms like ethics boards, red-teaming, and regulatory engagement.

Critics including Tristan Harris, co-founder of the Center for Humane Technology, have called the dynamic a risk of "ethics washing" or "pope washing," warning that visible alignment with religious moral authority does not resolve open technical questions about controlling increasingly capable AI systems. The relationship also sits alongside Anthropic's other entanglements: reported use of its technology in intelligence and defense contexts, an ongoing $1.5 billion copyright settlement, and friction with the Trump administration over AI regulation, including a $20 million Anthropic donation to a group backing regulation-friendly midterm candidates. Whether the Vatican's involvement shapes the regulatory debate the White House paused in May 2026, or whether other labs pursue similar faith-leader engagement, are the signals worth tracking next.