Gathered around a long conference table teeming with rainbow cookies and rugelach this past January, a group of Haredi rabbis in Lakewood, New Jersey, convened for a long discussion about the theological implications of artificial intelligence. (Haredim are Orthodox Jews with a strict interpretation of Jewish law and tradition.) The old men, attired in black-rimmed hats and sipping seltzer, planned a communal fast to protest the technology’s incursion into their communities, and discussed measures to eliminate its use in schools and homes. While this meeting was attended by elders of four Hasidic dynasties and other leaders in the Haredi world, traditionalist Orthodox Judaism lacks a central governing body, so the meeting’s outcome was by no means globally binding. Other Haredi groups even mocked the rabbis’ attempts at an AI ban, and argued that AI tools have kosher uses.

The episode has striking parallels with secular debates around AI. It’s worth thinking about why the discussion within a fervently traditionalist community played out so similarly to the one on the secular left. The Lakewood rabbis don’t consider generative AI destructive simply because it is new or because it could easily supply users with blasphemous content (though there is that). They also argue that interpreting the Torah is a fundamentally communal activity that requires ameilut, or toil, a philosophy espoused by foundational scholars like Rashi. They fear a future in which Torah study is fully automated and human intellectual exertion nullified. In this, they’re surprisingly aligned with leftists who oppose data centers and big tech. “They might come up with different sources for explaining it,” anthropologist Ayala Fader told The Forward, “but they are actually articulating some of the objections to AI that you can read about in The Chronicle of Higher Education.”

There are, of course, more prosaic concerns. Haredim, like other groups that enforce strict religious or cultural observances, have long argued about when and how to accommodate the demands of the contemporary world, and whether to adopt or shun new technologies. A decade ago, a debate raged among Hasidim (a subset of Haredim) about the Internet’s potential for sacrilege. Today, there are Hasidic TikTok influencers—although they abide by strictly enforced community standards. Many Orthodox Jews today see AI as a similar gateway to pornography or other impurities, and it remains to be seen whether they will accept it as inevitable and just try to put up some guardrails, or succeed in fully eradicating it. The Satmar Rebbe, the head of the Satmar Hasidic dynasty, put out an emergency broadcast this past May: “On every phone in the home, even on a kosher phone, one can connect to…impurity in a single moment,” he said. “There are many victims—young boys, girls, and even married yeshiva students. They are falling and descending directly into hell.” He called upon Satmar Hasidim to abstain from cell phones that preload applications like Gemini or OpenAI.

Anti-AI Haredim are not alone in considering the risks of this new technology to be existential. The threats it poses to labor are obvious, and yet there’s something about it that’s still more troubling, something that can be hard to put your finger on. The debates around AI often take on a strangely religious tone—after all, the most prominent leader to condemn its encroachment on humanity is Pope Leo. The Lakewood rabbis’ objections may help us better understand why both secular and religious AI refuseniks feel that something elemental is at stake. The Haredim who met in January argued that mitzvot—good deeds performed to fulfill a Torah commandment—cannot be replaced by machines. To eschew ameilut is to forgo the sacred work of Torah study. Outsourcing interpretation of scripture can be seen as an attack on the act of exegesis itself. In this way, AI threatens to undermine their way of life more thoroughly than previous technologies.

While it may be dangerous, AI is also increasingly omnipresent, complicating Orthodox communities’ attempts to rein it in. “You have to create rules that people will follow,” Forward staff reporter Louis Keene told me over the phone. “And if a rabbi is setting rules and nobody’s following them, it doesn’t matter how strict he is because nobody’s observing it.” Yet Keene acknowledges that many people intuit that there’s just something off about AI: “A lot of people are either nervous about it or repulsed by it. I don’t think people generally like AI art. Why do we see so much slop? We’re calling it slop. Everyone agrees.” For religious as well as secular communities, that slop presents a kind of spiritual affront.

There is a remarkably religious flavor to the way even atheists and technocrats talk about AI. Advocates frequently refer to it as a god in the making. “If we ever get AI that is strong enough to basically be God and solve all of our problems, it will need to use the same techniques that the actual God uses in terms of maintaining some distance,” rationalist and AI blogger Scott Alexander told Harper’s Magazine earlier this year. I do think it’s possible that the AI will be like, Now I am God. I’ve concluded that the actual God made exactly the right decision on how much evil to permit in the universe. Therefore I refuse to change anything.” Why are self-styled rationalists leaning so hard into such an esoteric framing?

“[T]here comes a time in every ai policy professional’s life when they realize they have to read the talmud to make further progress,” Dean W. Ball, OpenAI’s head of strategic futures, posted this summer. While that may seem like a completely out-of-left-field remark to anyone not versed in Silicon Valley lore, it sparked extensive conversations among AI researchers. The thinking goes that the Talmud offers a model of interpretation and debate that could teach AI the nature of ethics. Researchers often refer to the “AI alignment problem,” which is the difficulty that human designers face in making sure an AI system’s behavior matches the goals and values humans actually want. This is both a practical and philosophical problem, so loaded that it lends itself more easily to the domain of theology than corporate governance. Judd Rosenblatt, founder of the AI consulting company AE Studios and the AI ethics nonprofit AI Alignment Foundation, told The Jewish Forward that he looks to the Talmud for inspiration as he considers the AI alignment problem, believing its tradition of evolving arguments and adaptations could help us design AI systems that can improve exponentially without going rogue. “I think a tradition that reinterprets nothing is the more fragile one,” he said. “A rule that cannot be bent, cannot adapt to a new world and dies out.” Besides, he argues, Judaism has survived multiple implosions, including the destruction of the temple.

Not all Haredi agree that AI is a modern-day Tower of Babel, tempting the wrath of God with our hubris. Users on imamother.com, an online forum for Orthodox women, discuss their use of AI with a mix of shame and excitement, with some saying it motivates them to see doctors or go to the dentist. As one poster asked, “Am I a horrible person for not listening to das torah [a rabbinic pronouncement]? I can’t believe that HaShem would rather I be completely dysfunctional than use this tool.”

Other Orthodox Jews are embracing these tools wholeheartedly, particularly those identified with Modern Orthodoxy, a broad movement that seeks to reconcile Jewish observance with modern life. Orthodox Union, a large US-based nonprofit, is developing AI-powered Torah Study apps that collate large amounts of revered texts for both professional scholars and curious laypeople. Orthodox Union’s Rabbi Simon Taylor discussed the organization’s reasoning behind developing an app: “We have to figure out how to leverage AI to teach more Torah and to spread Judaism. We knew that we weren’t going to get it right the first time and it wasn’t going to be perfect. We worked very, very hard on safeguarding and guardrails for everything that we do as well, lots and lots of testing.” Rabbi Zohar Atkins has a new venture called Yochai, a chatbot that offers AI study support for the Torah and Talmud. The study tool takes its name from the second-century sage Shimon bar Yochai and its logo is a carob tree—the plant that, according to legend, the sage sustained himself on as he hid in a cave for 13 years to escape Roman persecution.

On the other hand, Rabbi Danya Ruttenberg recently delineated a Jewish case for being exempted from using AI in the workplace. She argues that AI is theft, as large language models like ChatGPT culled their data from countless sources without permission or compensation. Halakha, Jewish law, forbids Jews from benefiting from stolen goods. Similarly, claiming authorship for something one didn’t write oneself also a form of theft—something rabbis using AI to craft sermons may find alarming. She also points to the destruction that data centers are wreaking on the environment, and invokes the ethical principle of Bal taschit, which God reveals to Moses in Deuteronomy 20:19: Do not destroy.