AI Safety Donors Paid Religious NGOs $3.3M for Statements on AI

In 2024, Trump defeated Harris to succeed Biden as president. AI Safety’s largest donors — Dustin Moskovitz and the now-convicted fraudster Sam Bankman-Fried — were among Biden’s largest donors. Suddenly, they were on the outside looking in. Trump rejected the movement’s AI doomsday prophecies and repealed Biden’s Executive Order on AI.

With enough funding, they found a surprising solution — religious NGOs. Since 2025, AI Safety donors put $3.3 million into churches, seminaries, faith networks and religious-affinity groups. It paid for pronouncements: declarations, red lines, and a “Magna Carta of AI.”

The AI Safety movement descends from rationalist and new atheist roots. Its core belief — that AI will literally end all human life — often draws comparisons to doomsday cults like Heaven’s Gate. Yet with enough funding, some churches are willing to square that belief with the Book of Revelation.

Grant descriptions leave no doubt about what FLI paid for: public statements about AI.

The Greek Orthodox Archdiocese of America took $105,000 from FLI. They credit their “red lines on AI” to “a grant from the Future of Life Institute.”² The Gospel Coalition received $200,000 for a podcast called “Silicon Spiritualities,” without disclosing their sponsor. They later barred their writers from using AI to research, write, or edit. The World Council of Churches took $100,000 to write its “red lines.” The Mormon nonprofit Faith Matters took $299,000 for a “Magna Carta of AI.”

FLI, an AI Safety megadonor, runs a full-time religions program with a “Religions lead,” Will Jones, and a “U.S. Faith Liaison,” Brian J. A. Boyd. The program lists the Pontifical Academy of Sciences and the Vatican’s Fratelli Tutti Foundation among its partners.³

FLI’s grantees make up the overwhelming majority of religious signatories on statements produced by FLI and its allies.⁴

Of FLI’s recipients, only two disclosed their donations on their project pages. Two more disclosed their donations, but only on LinkedIn. Four more made no public project-related disclosures.

FLI’s other religious outreach is not progressing as smoothly. They paid a Muslim NGO for a declaration of Islamic principles and Impact Academy for a Hindu consensus statement. Neither has been delivered.

In September 2025, a Vatican working group chaired by papal adviser Fr. Paolo Benanti handed Pope Leo XIV an appeal for a prohibition on superintelligence “until there is broad scientific consensus that it will be done safely and controllably,” plus an IAEA-style inspectorate for AI labs.⁶ By November, FLI sat on the Vatican’s Builders AI Forum. In May 2026, it put $200,000 into two Catholic projects: a post-liberal magazine in Steubenville, Ohio, and a Slovenian faculty educating Catholic leaders on AI.

FLI is using its newfound connections to support the business disputes of Anthropic, an AI-safety-linked company. In March, Boyd co-signed a Catholic theologians’ brief backing Anthropic in its dispute with the Department of War. Anthropic brought him to the company’s two-day summit with fifteen Christian thinkers.⁷ Then, on 25 May, Pope Leo XIV presented Magnifica Humanitas, the first encyclical devoted to artificial intelligence, and the lay speaker beside two cardinals in the Synod Hall was Chris Olah, co-founder of Anthropic.

The AI Safety movement spends more brazenly and deceptively than almost any other political cause. This affects not just the AI debate, but theology broadly. As Nirit Weiss-Blatt puts it, “AI doomerism fills a ‘religious-shaped hole.’ For people who left religious upbringing but retain a hunger for a mission-driven community, the AI-risk subculture offers a replacement meaning system.”