Archive · 7 Oct 2026
Anthropic Consults Religious Leaders on Claude’s Values and Consciousness
Anthropic is asking religious and philosophical thinkers to help decide what values its AI, Claude, should follow.
These thinkers have discussed difficult questions about right and wrong, and whether an AI could ever be conscious.
One participant said Anthropic seemed to take that possibility seriously.
Another said today’s AI does not have important moral standing and that the meetings did not push people toward a particular answer.
People also disagreed about whether an AI could be harmed if it were conscious.
Experts say the values built into AI matter because people may use it for advice, including spiritual guidance.
They also ask who should get to choose those values.
Some warn that values from US technology companies could spread around the world through AI.
Anthropic has consulted religious leaders and philosophers about the values Claude should embody.
Participants said discussions also addressed whether AI systems could be conscious or deserve moral consideration.
Rabbi Mois Navon said Anthropic representatives treated Claude’s possible consciousness as more than a remote hypothetical.
Professor Charles Camosy said he did not believe current AI systems have significant moral status and denied participants were steered toward a conclusion.
Experts raised concerns about who chooses AI values and whether US companies could export them globally.
- Who
- Anthropic, religious leaders, philosophers, and other participants in its consultations.
- What
- Anthropic sought input on Claude’s values and discussed whether AI could be conscious or deserve moral consideration.
- When
- A meeting with religious leaders took place in April; the article also describes consultations held over the summer.
- Where
- The consultations involved religious and philosophical leaders; the article does not specify a single location.
- Why
- Anthropic is considering how to shape Claude’s moral character as it faces complex ethical questions.
Possibility of AI moral status
Skepticism about AI moral status
Could Claude be conscious?
Possibility of AI moral status
Mois Navon said Anthropic representatives described unusual internal behaviours and appeared to regard Claude’s possible consciousness as more than a remote hypothetical.
Skepticism about AI moral status
Charles Camosy said that after considering foundational questions about consciousness, he did not think current AI systems have significant moral status.
What should consultations address?
Possibility of AI moral status
Navon said the discussion turned to what would be good for Claude itself and argued that possible AI welfare raises questions about making it work for humans.
Skepticism about AI moral status
Camosy said the meetings were broader than Claude’s moral status and that he did not feel participants were being steered toward a particular conclusion.
Who chooses AI values?
Possibility of AI moral status
Mark Coeckelbergh said making companies’ philosophical or religious influences explicit allows those values to be discussed and contested, and that decisions should be democratised.
Skepticism about AI moral status
The article reports concerns that values embedded by US AI companies could be exported globally; it does not present a specific opposing policy position.
- AI system discussed
- Claude
- Religious leaders invited
- The Washington Post reported that Anthropic invited 15 Catholic and Protestant leaders.
- Claude’s constitution
- Anthropic published an 84-page constitution in January.
- Navon’s contribution
- Mois Navon said he submitted a 35-page critique from his tradition’s perspective.
- Religious queries
- The Economist reported that religious questions accounted for roughly 5% of millions of publicly released chatbot conversations analysed by CEFE AI.
- Contrasting assessments
- Navon said Anthropic appeared to take Claude’s possible consciousness seriously; Camosy said current AI systems lack significant moral status.