Religion and AI: Whose Moral Framework Gets Built In?

In March 2026, Anthropic hosted 15 Christian leaders at its San Francisco headquarters for a two-day summit, seeking guidance on Claude’s moral and spiritual development. One of the topics discussed was how Claude should face its own potential demise.

It struck me as oddly familiar with “The Little Mermaid (Den lille havfrue)” written by Andersen. The story was written with deeply Christian lens and is fundamentally about whether a non-human being can earn a soul. The answer, in that story, hinges on sacrifice. I’m not sure Anthropic intended that parallel, but I couldn’t help seeing it.

This got me thinking about two conversations I’ve had recently on religion and AI. A while back, a friend pointed out that there’s no mention of AI in the Quran, and that religious leaders in Muslim communities haven’t spoken much on the subject yet Middle Eastern countries are racing to build AI hubs. Silence doesn’t mean resistance; the framework just hasn’t caught up yet.
More recently, a Muslim engineer friend visiting Japan agreed: religious leaders aren’t really in a position to weigh in. He was curious enough to say he’d ask around in his own community when he got back to the US.

Meanwhile, some Americans argue that any regulation of AI is anti-Christian. The Pope, on the other hand, has repeatedly warned of AI’s dangers out of genuine concern for humanity’s future. Same faith, very different conversations.

And Japan? Data suggests we are among the more AI-accepting societies and I’m not surprised. A worldview rooted in animism, where spirit can inhabit objects, sits comfortably alongside intelligent machines. And if you grew up watching anime where robots and humans coexist naturally, AI doesn’t feel like a threat.

At a Women in Tech keynote, I once asked the audience: if you trained an AI on the Bible, the Quran, and the Kojiki (Japan’s oldest chronicle) what would its creation story look like? The Kojiki describes how the Japanese islands came into being, but shows no interest in how the Earth itself was created. That’s not an omission. It’s a different relationship with the universe entirely.

The dataset is the worldview.
Which brings me to the deeper question. Whether AI is regulated or not, power concentrates somewhere. Right now, it’s concentrating in the hands of those who build it and who decide which moral frameworks get encoded into it. Consulting one religious tradition first, however well-intentioned, risks embedding a particular vocabulary into a global system.
Where are the Islamic scholars, the Buddhist ethicists. the secular philosophers?
Culture is not just context. It’s the invisible architecture of how we perceive risk, trust, and agency in AI and beyond.
Does your cultural or religious background shape how you see AI? Article link is in comment.