Anthropic Enlists Faith Leaders to Shape AI Morality

Anthropic, the company behind the Claude AI model, is taking an unconventional approach to building its moral compass. Rather than relying on a single philosophical or scientific framework, Anthropic is seeking input from religious thinkers around the world.

The company's interest in faith-based perspectives began with a 2025 Vatican note on artificial intelligence, titled 'Antiqua et Nova.'

Anthropic has since convened meetings and consultations with leaders from various Christian denominations, including Catholics. The most notable example came in March 2026, when the company brought together 15 Christian leaders to discuss AI ethics.

The goal is for Claude to act as a 'good, wise, and virtuous agent,' according to Anthropic's Constitution. This document serves as the model's written set of values and guidelines, which will help it reason through complex requests.

The company's approach avoids locking Claude into a single ethical doctrine, instead drawing on diverse perspectives to create a more nuanced moral framework.