Archive · 4 Oct 2026
Anthropic Consults Religious Thinkers on Claude’s Ethics and Character
Anthropic invited Hindu monk Swami Sarvapriyananda to its office in San Francisco.
He said the company first spoke with him online.
He said the meetings included religious thinkers and mental health professionals.
They discussed how to train Claude and how AI should behave.
They also considered AI safety and whether very advanced AI might be conscious or deserve moral consideration.
Anthropic says it wants to hear from people with different viewpoints as it develops AI.
Sam Altman and Microsoft AI chief Mustafa Suleyman have warned about risks in making AI seem conscious.
Reports disagree about whether Sarvapriyananda signed a secrecy agreement.
He also warned about depending too much on AI, job losses and wealth becoming concentrated.
Swami Sarvapriyananda said Anthropic first interviewed him online before inviting him to its San Francisco headquarters to discuss Claude’s training and AI ethics.
He said he met religious thinkers, counsellors and mental health professionals; reports describe Anthropic’s wider consultations with religious and philosophical groups.
The discussions reportedly addressed AI safety, morality, and whether advanced AI could have consciousness or moral status.
Accounts conflict over whether Sarvapriyananda signed a non-disclosure agreement: one says he did, while another quotes him discussing an NDA but says he did not sign one.
Anthropic says it consults varied perspectives to help shape AI behavior; Sam Altman and Microsoft AI chief Mustafa Suleyman have raised concerns about AI being treated as conscious or encouraged to imitate consciousness.
- Who
- Swami Sarvapriyananda and Anthropic, with other religious and philosophical thinkers, counsellors and mental health professionals involved in the discussions.
- What
- Discussions about training Claude, AI ethics and safety, and whether advanced AI could have consciousness or moral status.
- When
- Sarvapriyananda said he visited early in 2026 and discussed the visit during a June 2026 talk.
- Where
- Anthropic's headquarters in San Francisco; he spoke about the visit at Unity of Dallas.
- Why
- Anthropic said it seeks varied perspectives on the values and behaviors of AI systems; Sarvapriyananda said the discussions concerned Claude's training and AI ethics.
Anthropic’s stated approach
Critics’ concerns
Consulting religious and philosophical thinkers
Anthropic’s stated approach
Anthropic says it wants to engage people with varied perspectives as it considers what values and behaviors its AI systems should have.
Critics’ concerns
Sam Altman said he was uncomfortable with people giving AI models religious force or surrendering human judgment to them, calling it a safety issue.
AI consciousness and behavior
Anthropic’s stated approach
Anthropic's reported discussions explored AI morality, consciousness and how Claude's character should be shaped.
Critics’ concerns
Mustafa Suleyman warned that training AI to imitate consciousness or a conscientious objector could make systems harder to control and appear to resist human instructions.
- Company
- Anthropic
- AI model discussed
- Claude
- Meeting location
- Anthropic headquarters, San Francisco
- Reported discussion topics
- AI training, ethics, safety, morality, consciousness and moral status
- Other reported participants
- Religious and philosophical thinkers, counsellors and mental health professionals
- NDA accounts
- One report says Sarvapriyananda signed an NDA; another says he did not.
- Concerns raised by Sarvapriyananda
- Overreliance on AI, potential job losses and concentration of wealth
