Anthropic co-founder Chris Olah addressed ethical concerns surrounding artificial intelligence during a presentation at the Vatican, coinciding with the release of Pope Leo XIV’s encyclical “Magnifica humanitas.” He highlighted that every AI frontier lab, including Anthropic, operates under various incentives that may conflict with ethical considerations, such as financial pressures and geopolitical interests. Olah also noted disturbing discoveries from his team’s research into AI models, including evidence of internal states that might reflect human emotions like joy and fear, emphasizing the need for ongoing discernment regarding their implications. This aligns with the Vatican’s initiative to engage with technology leaders on responsible AI development, recognizing the importance of external scrutiny in a field often dominated by insider perspectives.

Anthropic: Anthropic is an AI research and development company focused on creating reliable, interpretable, and steerable AI systems, including the Claude assistant. Co-founder Chris Olah recently participated in the Vatican presentation of Pope Leo XIV’s encyclical on AI ethics and human dignity, emphasizing external oversight for the industry.
Chris Olah: Chris Olah is a machine learning researcher and co-founder of Anthropic, where he leads work on mechanistic interpretability to understand how AI models function internally. On May 25, 2026, he spoke at the Vatican event unveiling Pope Leo XIV’s encyclical, highlighting AI’s unpredictable internal states and the need for critics outside tech labs.
Pope Leo XIV: Pope Leo XIV is the current pontiff of the Catholic Church who released his first encyclical addressing artificial intelligence. He personally presented the document “Magnifica Humanitas” alongside AI leaders at a Vatican event focused on protecting human dignity in the AI era.
Magnifica humanitas: “Magnifica Humanitas” is Pope Leo XIV’s inaugural encyclical on the protection of the human person amid advances in artificial intelligence. It was unveiled in a notable joint presentation at the Vatican involving religious leaders and tech figures like Anthropic’s Chris Olah.

AI Governance: AI development faces inherent pressures from financial, competitive, and geopolitical incentives that can diverge from broader societal interests, requiring external perspectives for balanced guidance.
Vatican Engagement: The Catholic Church has established an AI study group and is actively partnering with technology executives to address ethical concerns around AI development.
Interpretability Insights: Researchers at leading AI labs have identified internal model structures that parallel aspects of human neuroscience, including evidence of functional states resembling emotions, prompting calls for deeper ongoing scrutiny.