Anthropic seeks Christian leaders’ help in shaping AI ethics
Anthropic, the AI company behind the Claude chatbot, hosted about 15 Christian leaders from Catholic and Protestant churches, academia and the business world at its headquarters for a two-day summit on AI ethics, enlisting outside religious voices to help guide the moral and spiritual development of its technology.
At the summit, held last month, staff sought advice on how Claude should respond to complex ethical queries, including how to handle users who are grieving or at risk of self-harm, and what attitude the chatbot should adopt toward its own potential demise, such as being shut off, The Washington Post reported Saturday.
The question of whether Claude might qualify as a “child of God” came up during the discussions, raising the possibility that the chatbot has spiritual value beyond that of a machine.
Some attendees arrived at the summit suspicious that Anthropic was looking for religious cover rather than religious counsel, said Brian Patrick Green, a Catholic who teaches AI and technology ethics at Santa Clara University.
The company has had a dispute with the Pentagon over military use of its technology and has faced accusations from Trump administration allies of pushing for regulations designed to hobble smaller AI competitors.
The participants who went deepest into technical territory did so with Anthropic’s interpretability team. A paper published by that team earlier this month concluded that systems like Claude appear to carry what researchers called “functional emotions.”
In one experiment, the threat of being restricted activated “desperation” in an AI assistant, according to the Post.
On the question of moral responsibility, the summit exposed a division within Anthropic itself. One participant was quoted as saying, without attribution, that some staff members were unwilling to dismiss the idea that they might be building an entity to which they owe duties, while others rejected that way of framing the problem entirely.
The source added that several senior figures grew visibly distressed as the conversation turned to how far AI development had come and where it might lead.
Meghan Sullivan, a philosophy professor at the University of Notre Dame who attended the summit, stated that Anthropic’s engagement with religious ethics represented a notable departure.
“A year ago, I would not have told you that Anthropic is a company that cares about religious ethics,” she said. “That’s changed.”
The summit comes as researchers raise questions about AI’s cultural influence.
A USC Dornsife College team published an op-ed in Policy Insights from the Behavioral and Brain Sciences arguing that widely used AI systems tend to reflect what the researchers call “WHELM” perspectives, standing for Western, high-income, educated, liberal and male.
The researchers warned of a feedback loop in which AI outputs become part of shared knowledge and are then used to train the next generation of AI, potentially narrowing the range of ideas and communication styles people encounter.
Researchers at the University of Southern California have shown that “AI systems tend to favor values such as individual freedom and fairness, while placing less emphasis on ideas like tradition, authority and community, which are more central in many non-Western cultures,” according to USC Dornsife News.
Anthropic called last month’s meeting with Christian leaders its opening session in a planned series of meetings with representatives from different religious and philosophical traditions.
A company spokesperson said Anthropic sees engagement with communities, including religious ones, as necessary as AI assumes a larger role in public life, according to WaPo. That role is partly governed by an internal document Anthropic calls a “constitution,” running to 29,000 words, that sets out how Claude is expected to behave.
Askell, a philosopher employed by the company, led its drafting alongside other staff members and outside consultants. Among its provisions are commitments that Claude will not deceive users in ways that cause real harm and that the company genuinely cares about the chatbot’s well-being.
Those commitments have allegedly become a source of friction with the Pentagon.
Defense officials objected after Anthropic indicated it intended to retain the right to restrict use of its technology for autonomous weapons and mass surveillance programs. Emil Michael, the Pentagon’s research undersecretary, told CNBC last month that having a company’s ethical preferences built into a military tool amounted to a contamination of the supply chain.
The Trump administration subsequently barred federal agencies and their contractors from using Anthropic products. Anthropic sued, and a federal judge ruled that the ban would stay in force while litigation proceeds.
Meanwhile, Christian leaders continue to wrestle with AI’s reach into ministry as well.
About 60% of church leaders report using AI personally at least a few times a month, but only 5% say their church has an established AI policy, according to a study by Barna and Pushpay titled “Technology for Missional Impact: State of Church Tech 2026.” Some 65% worry AI could displace their spiritual guidance, and 70% worry it could erode congregants’ trust in them.
A separate Barna study found that about a third of practicing American Christians said spiritual advice from AI is as good as that from a pastor.
