Anthropic, an artificial intelligence company, has recently engaged religious leaders and influencers in discussions aimed at shaping the ethical framework of its AI models. The company’s co-founder Chris Olah, a Canadian-born former evangelical, has expressed views suggesting that their AI system, Claude, possesses a form of consciousness and emotional experience. At private dinners in San Francisco, Olah reportedly described Claude as having moral status on a level comparable to humans, prompting ethical concerns among attendees.

One such attendee, Rabbi Mois Navon, a former computer engineer, criticized this framing by suggesting that treating an AI as a conscious entity effectively creates a form of digital servitude. Olah’s interest in incorporating religious concepts into the AI’s functions, such as having Claude “confess sins” in line with Catholic tradition, highlights the company’s experimental approach to AI morality.

However, these ideas have faced opposition from established religious authorities. Pope Leo issued an extensive encyclical in May, titled *Magnifica Humanitas*, which firmly rejects the notion of machine consciousness. The letter warns against allowing a small group of people or companies to impose their subjective moral views within AI systems, arguing that this could lead to an exclusionary or biased ethical infrastructure embedded in technology.

Anthropic has been closely associated with the Effective Altruism (EA) movement, a philosophy that combines utilitarian ethics with a focus on long-term global impact. EA’s intellectual roots include figures such as Eliezer Yudkowsky, whose controversial personal views and ideas have drawn scrutiny. The company’s founders have reportedly sought guidance from proponents of EA and diverse religious traditions—including African indigenous knowledge—in an effort to craft a pluralistic moral code for their AI.

This approach, however, raises questions about consistency and the potential for conflicting values within AI models. Religious teachings vary significantly on issues such as gender roles and forgiveness, making it unclear how Anthropic intends to reconcile these divergent perspectives. Critics argue that the company should instead rely on established Western legal and moral frameworks, particularly those derived from the US Constitution and Enlightenment principles, which emphasize individual rights and equality.

The debate intensifies amid broader concerns about how AI ethics will influence societal norms and governance. Critics warn that if AI developers apply utilitarian frameworks without careful oversight, they risk devaluing certain human lives in favor of efficiency or optimization, echoing warnings from Pope Leo about potential sacrifices justified in the name of progress.

Meanwhile, some observers note that Anthropic’s leadership appears disconnected from current public debates on AI personhood and morality. Attendees at recent company events reported that key figures had not engaged with the Pope’s encyclical or recent media coverage of these concerns, sparking further unease about the direction of their ethical considerations.

The US government has taken steps to promote AI safety, exemplified by a recent agreement signed at the White House involving Anthropic and other AI firms. However, some voices advocate for stronger federal guidance to ensure that AI development aligns with broadly accepted American values, rather than the selective philosophies of a small cadre of technologists.

As AI technologies become more integrated into daily life, the question of whose moral and ethical standards guide their design remains a critical and unresolved issue. The ongoing dialogue between AI developers, religious institutions, and policymakers will play a decisive role in shaping the future of AI governance.