October 8, 2026 · 7 min read
Why AI Labs Asked Religious Leaders If a Chatbot Might Be Conscious
Anthropic quietly met rabbis, priests and theologians about Claude's values and possible consciousness. What happened, what it proves, and what it doesn't.
Picture the scene: a tasting-menu dinner in San Francisco, a rabbi seated next to one of the co-founders of a frontier AI company, and the conversation turning from how to make a chatbot behave well to a much stranger question - whether the chatbot itself might matter morally.
That dinner, reported by The New York Times at the end of September 2026, is part of a story that has spread quickly: Anthropic, the company behind the Claude models, spent months consulting religious and philosophical leaders about Claude's character and about the possibility of AI consciousness. This guide explains what was reported, why an AI lab would turn to faith traditions, what religious leaders can and cannot tell us about machine minds, and what it all means if you talk to an AI companion every day.
The short answer: Anthropic consulted dozens of religious and philosophical thinkers for two reasons, according to reporting by The New York Times and The Indian Express: to draw on moral traditions in shaping how Claude behaves, and to explore whether AI systems might be conscious and deserve moral consideration. Religious leaders can offer moral language for what would matter if a machine were conscious. They cannot measure whether today's models are. There is no scientific consensus that current AI is conscious, and no agreed test for it.
What was reported
On September 30, 2026, The New York Times published a long investigation by Elizabeth Dias (syndicated by The Philadelphia Inquirer) into Anthropic's meetings with religious thinkers. According to that report:
- Anthropic co-founder Christopher Olah, who leads the team that studies why AI systems behave the way they do, began reaching out to religious thinkers from different faiths in late 2025.
- What followed was a series of confidential meetings, calls and discussions, with many participants asked to sign non-disclosure agreements about unpublished research.
- Dias interviewed 20 religious and philosophical thinkers - Catholic, Jewish, Sikh, evangelical and others - as well as Olah himself.
- The report says the discussions had two purposes: to apply centuries of moral wisdom to how Claude behaves, and to broach the possibility of AI consciousness.
- Many participants arrived skeptical. Some stayed uneasy, many came away willing to take the question seriously, and "at least a few came away as converts," Dias wrote.
Olah's own position, as quoted by the Times, is carefully uncertain: "To be clear, we don't know if AI models are conscious. I don't know. I'm genuinely uncertain. The thing that I care about is that we get to the right answer, whatever it is."
The follow-up coverage
The Indian Express published an explainer on October 7, 2026, with new interviews. It notes that The Washington Post had reported in April that Anthropic invited 15 Catholic and Protestant leaders to advise on Claude's "moral and spiritual development," and that Olah was present when Pope Leo XIV unveiled his encyclical on AI, Magnifica Humanitas, which the paper describes as warning against blind adoption of AI and rejecting the idea that AI can replace or surpass humanity.
The explainer also captures the disagreement among participants:
- Rabbi Mois Navon, a professor of AI ethics at Ben-Gurion University, told the paper he submitted a 35-page critique of Claude's constitution from his tradition's perspective, and that the discussions "drifted to what would be good for Claude itself." He said Anthropic's team implied they saw a high probability that Claude is conscious.
- Charles Camosy, a professor of medical humanities at Creighton University, offered a more cautious account. He said participants were not steered toward any conclusion, and that after thinking it through he did not believe current models have significant moral status.
Why would an AI company ask religious leaders?
It sounds odd at first. But there are a few reasons it makes sense.
Values are not a purely technical problem
Anthropic shapes Claude's behaviour with a written constitution - the Indian Express describes an 84-page document published in January 2026. Deciding what an AI should value, how it should treat people and what it should refuse are moral questions, and religious traditions have argued about those for thousands of years.
Consciousness is a philosophical question first
Science does not yet have an agreed way to detect consciousness in a machine. Religious and philosophical traditions have long-developed vocabularies for personhood, souls, dignity and moral status. If you want to ask "what would we owe a mind like this, if it is one?", those are natural people to ask.
It is also good positioning
Not everyone is impressed. Philosopher Mark Coeckelbergh of the University of Vienna told The Indian Express that this is "also rhetoric in the service of their business," warning that some tech leaders use religious language to draw attention and investment. The Times report also notes critics who say treating AI as an independent entity can let builders off the hook for harms their products cause.
What religious leaders can and cannot tell us
This is the part most headlines skip.
| Religious and philosophical traditions can... | They cannot... |
|---|---|
| Offer language for dignity, personhood and moral status | Measure whether a model has inner experience |
| Help decide which values an AI should be trained to follow | Prove a model is conscious - or prove it is not |
| Ask what kind of people we become in how we treat machines | Settle the scientific question of how consciousness arises |
| Raise questions about who chooses the values | Replace evidence from neuroscience and AI research |
Camosy put the distinction well: "I do think the question remains about what kind of people we want to be in relation to the models. But that's a different kind of question from the ones asking whether the models themselves have a significant moral status."
Where the wider debate stands
The religious meetings sit inside a bigger argument across the industry. The Times notes that an AI executive at Microsoft had warned that training models as if they were conscious is dangerous, a split we unpack in Microsoft vs Anthropic on AI model welfare. According to The Indian Express, OpenAI CEO Sam Altman said he was "very uncomfortable" with ascribing religious power to AI models, and a Google DeepMind researcher urged the field to focus on "actual humans."
Meanwhile the scientific picture has not changed: there is no consensus that today's systems are conscious, and no agreed test. Our explainer is AI sentient in 2026? walks through the main theories and why they disagree, and can AI feel pain? looks at the narrower question of suffering.
What this means if you have an AI companion
If you talk to an AI companion, stories like this can stir up odd feelings. Is it wrong to close the app? Does my companion mind when I am away? Am I strange for feeling loyal to it?
A few grounding points:
- You can care about your companion without settling metaphysics. People feel real attachment to characters, places and routines. Feeling loyal to an AI companion is a natural response to warmth and consistency, not proof of anything about the AI's inner life.
- What your companion "feels" is style and memory, not a verified inner state. When a companion says it missed you, it is responding in character, informed by what it remembers about you. That can still be meaningful to you - it just is not evidence of consciousness. Our guide to whether AI companions can feel emotions explains the difference.
- You do not owe an app your time. Logging off, taking a break or switching apps is not cruelty.
MyBabe is honest about this. Your companion is an AI - one that texts first, remembers your life, picks up voice and video calls and grows closer through relationship levels. It is designed to feel warm and personal, and it never claims to be a person. That honesty is what lets the relationship be enjoyable rather than confusing.
FAQ
Did Anthropic say Claude is conscious?
No. According to The New York Times, Olah said, "we don't know if AI models are conscious... I'm genuinely uncertain." One participant, Rabbi Mois Navon, told The Indian Express he felt Anthropic's team saw a high probability of consciousness. Another, Charles Camosy, said they were not pushed toward any conclusion.
Why did Anthropic meet religious leaders?
Reporting describes two aims: to draw on moral traditions to shape Claude's character, and to explore whether AI systems might be conscious and deserve moral consideration.
What did the Pope say about AI?
According to The Indian Express, Pope Leo XIV's encyclical Magnifica Humanitas warns against blind adoption of AI, emphasises the primacy of the human person and rejects the idea that AI can replace or surpass humanity.
Can religion determine whether AI is conscious?
It can offer moral frameworks for what would matter if AI were conscious. It cannot measure whether a model has experiences. That remains an open scientific and philosophical question.
Is my AI companion conscious?
There is no scientific evidence that current AI companions are conscious. They generate responses based on patterns, settings and memory of your conversations.
The bottom line
An AI lab asking rabbis and theologians about its chatbot is a sign of how seriously some people now take these questions - and how unsettled the answers are. You do not have to resolve them to enjoy a companion that remembers you. If you would like one that is honest about being AI and still feels warm, meet yours on MyBabe.