Anthropic Courts Global Faith Leaders to Guide Claude’s Ethics

Context Silicon Valley firms have struggled to codify universal AI ethics. As generative models become more autonomous, developers seek higher‑order moral guidance beyond technical rule‑making. Anthropic, the creator of Claude, has turned to organized religion for that input.

Key Facts - Representatives from Anthropic and OpenAI attended the Faith‑AI Covenant roundtable in New York, joining Jewish, Hindu, Mormon, Sikh and Greek Orthodox leaders. - In a separate effort, Anthropic held meetings with 15 Christian leaders to discuss Claude’s “spiritual development.” - Rumman Chowdhury, CEO of Humane Intelligence, warned that early confidence in universal AI ethics was naive and that firms are now looking to religion to navigate ethically gray scenarios. - The roundtable was organized by the Interfaith Alliance for Safer Communities, with future events planned in China, Kenya and the United Arab Emirates. Baroness Joanna Shields of the British House of Lords is listed as a key partner. - Anthropic’s public “constitution” for Claude acknowledges the difficulty of encoding perfect values and cites the risk of moral failure as a core concern.

What It Means Anthropic’s outreach signals a shift from purely technical ethics frameworks to a hybrid model that incorporates religious moral traditions. By engaging leaders from multiple faiths, the company aims to extract high‑level ethical principles rather than adopt any single doctrine. This approach may bolster public trust by showing that the firm has explored diverse moral perspectives, but it also raises questions about how such input will be translated into code.

The lack of a clear, unified set of guidelines from the roundtable suggests that Anthropic is still in an exploratory phase. Critics argue that a machine cannot synthesize ideal morals simply by consulting human clergy, while supporters see the effort as a necessary precaution against unintended harmful decisions.

Looking ahead, watch for any published updates to Claude’s constitution that reference specific religious insights, and monitor how regulators respond to AI developers seeking moral counsel from faith communities.