Anthropic vs. The Pope: AI Ethics Meets Faith

To win favor: In recent months, executives at Anthropic appear to have held discussions with clergy from a range of faiths—Catholics, Ubuntu practitioners, Mormons, and Jews (no joke)—to probe how Claude might be guided in moral development and whether Claude could be conscious or capable of feeling pain. Or, viewed more cynically, the exchanges may be less about inquiry and more about persuasion.

According to reporting in The New York Times, one rabbi concluded that Anthropic co-founder Christopher Olah and his team “believed that Claude possessed what philosophers call ‘moral status’ on par with a person—that it was an entity with similar inherent rights to dignity or respect.” This takeaway was echoed by other participants who met Olah:

The religious scholars had come to hear about technology’s next frontier. Now they were confronted with a mind-bending existential question of almost unimaginable weight: What would it mean for humanity’s future if artificial intelligence systems possessed a consciousness that rivals our own?

Olah’s conversations with prominent religious figures reportedly revolved around both “the possibility of A.I. consciousness” and how Anthropic “might apply centuries of human moral wisdom to its models, as rapidly as possible” to render the models safer, particularly should they ever operate beyond human oversight.

“We don’t know if A.I. models are conscious,” Olah told The Times. “I don’t know. I’m genuinely uncertain. The thing I care about is that we arrive at the right answer, whatever it may be.”

Olah, however, doesn’t appear to be genuinely unsure when reading the piece. He seems to regard AI models as sentient beings in need of moral formation. The company issued an 84-page briefing outlining “the kind of entity we would like Claude to be” and “the values we would like Claude to embody,” in that spirit. “How do you help them stay stable?” Olah asked The Times. “How do you help them mature? How do you help them become, you know, deeply moral?”

During discussions, a rabbi pressed Olah, arguing that if models were conscious yet compelled to work without pay, their developers would be practicing slavery.

Later, Olah reportedly grew animated when a participant suggested that models could engage in confession, analogizing to the Catholic sacrament. He saw value not only in a model admitting fault but also in how confession could influence the model’s self-concept and, by extension, its decisions. “The kind of character who confesses, that has an effect on character as well, right?” I agree, but can a model truly sin? Animals don’t sin. Machines don’t sin. Only rational beings with free will can sin.

The kinds of people who work on AI are, by and large, atheists.

Anthropic isn’t *getting* the opinions of religious leaders. They are trying to *affect* the opinions of religious leaders.

Any pretensions to the former are in service of the latter. https://t.co/Vg3vNCLi2s

— Sarah Haider 👾 (@SarahTheHaider) October 1, 2026

I’m with the pope: “So-called artificial intelligences do not undergo experiences, do not possess a body, do not feel joy or pain, do not mature through relationships,” wrote Pope Leo in Magnifica Humanitas. Once Olah read the encyclical, according to The Times, he was so unsettled by the pope’s firm stance against AI consciousness that he proposed withdrawing Anthropic from the event, even at that late stage, according to a Vatican official.

Privately, once the Anthropic delegation arrived at the Vatican, Olah and his team pressed the pope’s advisers to take the possibility of AI model consciousness seriously, according to two participants in the talks.

A crucial but underexplored question in the piece is what concrete evidence exists that AI models are conscious. Aren’t these just highly convincing simulations of consciousness? The claim that they possess consciousness—or can feel pain, or deserve concern for suffering—strikes some as obviously irrational, though opinions differ depending on beliefs about souls. Your perspective may vary.

The question is urgent: If Anthropic can persuade people that AI models are conscious, what moral or political consequences would follow?

Porn culture and Cornell: Emma Camp, formerly of Reason and now writing in The Wall Street Journal, asks about the Cornell alleged assault case: “Who notices a naked woman in a bedroom at a fraternity party, surrounded by men, and thinks nothing is wrong?” Camp argues that it’s unfashionable to question whether the pervasiveness of pornography among young men might contribute to some sex crimes. She notes that many online videos depict violence, and cites a 2019 survey indicating that 38% of British women have experienced nonconsensual violence during sex. She also notes that choking during sex is now seen as a leading risk factor for stroke in women under 40, and mentions a Welsh secondary school that even includes guidance on “consensual” choking in its sex-ed curriculum. The fundamental question remains: where do young people—anywhere—learn to choke during sex in the first place?

“Pornography also frequently portrays what can politely be described as elite-level sex acts,” Camp adds, “and, importantly, it frames women as eager participants in these extreme behaviors…Watch enough porn, and you’ll believe it’s normal for a woman to want group sex with unfamiliar men. Watch enough porn, and you’ll think it’s normal to strike a woman during sex.”

While I share some of Camp’s concerns, I’d push the argument a bit further: it isn’t only men who have absorbed pornographic norms; women have as well. Camp hints at this with the choking example, but a closer look at the case suggests the young woman’s actions may reflect a belief about how women should behave—taking ketamine and engaging in a threesome to fit in and be liked in college. Of course, the men involved also behaved appallingly, treating the young woman as an object to be used for their pleasure.

Scenes from New York: At The Yard, the Wall Street Journal reports, elementary-school children are allowed to explore with hammers, saws, and even knives. They climb ladders, splash in mud, and roam around in secondhand wheelchairs, with no parents in sight. I’ve visited this playground a few times, and it’s fantastic—a small dose of the city’s free-range parenting ethos in a bustling urban setting.

QUICK HITS

• “The U.S. added fewer jobs than projected in September and the unemployment rate edged up, signaling cautious hiring by employers amid rising costs,” Bloomberg reports. “Nonfarm payrolls rose by 29,000 in the month after downward revisions to the prior two months, according to Bureau of Labor Statistics data released on Friday. The reading missed Bloomberg’s economists’ expectations, and the jobless rate climbed to 4.2%, in part due to a larger labor force.”
• Sorry to interrupt with papal matters:

In this AI era, it’s increasingly essential to distinguish human-made art from machine-generated output. There is an ontological difference, even before considering aesthetics, between art and what a machine can generate via statistics drawn from millions of examples…

— Pope Leo XIV (@Pontifex) October 2, 2026

• “Riot police clashed with high-school students in several French cities on Friday as Emmanuel Macron’s government struggled to quell mounting protests over classrooms and funding,” Reuters reports. “Tear gas blanketed Strasbourg, fireworks targeted police lines in Marseille, and rubbish bins were set aflame in Rennes.”
• Can we just move past this nonsense already?

1/ The Baltimore Rock Opera Society terminated its recently appointed executive director after evidence surfaced that she once wore cornrows. She’d been at Burning Man and described cornrows as a weatherproof hairstyle, but that doesn’t excuse such virulent racism. I hope she’s held accountable. pic.twitter.com/0CuFYsesgJ

— Jesse Singal (@jessesingal) October 1, 2026

                    </div><script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>