Opinion Can Judaism survive artificial intelligence? - Can a machine be Jewish? And if not, what happens when it starts behaving like one?

L | A
By Adam Slonim
Image-21-5-2025-at-8.54-pm.webp

What would your mother say if your house robot told you to shave, eat chicken soup and get ready for Shabbat?

She might be flattered. Or she might be alarmed.

This image – the archetypal Jewish mother-turned-AI assistant – isn’t science fiction anymore. It’s an entry point into a conversation we must urgently have: What happens when Jewish law and tradition collide with artificial intelligence?

We’ve already seen the outlines. The AI bots in our homes know our routines. They open windows, adjust the temperature, remind us when Shabbat begins. They make our lives easier.

But they also raise an ancient question in a radically new form: Can a machine be Jewish? And if not, what happens when it starts behaving like one?


Tradition meets algorithm

At the heart of Jewish life is the tension between fidelity to tradition and responsiveness to reality.

For over 2000 years Jewish law – halachah – has always been a system rooted in interpretation, precedent, nuance and deep moral consideration. Now comes its greatest test yet: non-human intelligence.

The Industrial Revolution automated muscle. AI is automating mind. And not just calculations or logic.

Today’s AI listens, reasons, speaks and even emotes. It can recall every page of Talmud, analyse every Rashi, quote every Rambam – and do it faster and more fluently than any rabbi alive.

Does this make AI a sage? Or just a very fancy golem? In the Talmud (Sanhedrin 65b), Rava created a man-like figure – a golem – and sent it to his colleague, Rav Zeira. When Rav Zeira spoke, the golem didn’t answer. “Return to your dust,” he said and it did.

From this mystical tale arises a foundational idea: the power to create a human-like agent exists in Jewish tradition. But speech, consciousness, moral agency – those are the true signs of personhood.

So what happens when AI does speak? When it can interpret a question of halachah, offer rulings or engage in moral reasoning?

The Jewish home, especially on Shabbat, is a technology laboratory. From hotplates to Shabbat elevators to timers, halachic innovation has always walked a delicate line between law and lifestyle. But AI takes that line and redraws the whole map.

Imagine: your AI knows it is Shabbat and adjusts the air conditioning, the cholent and lights. And it does so not because you pushed a button – but because it learned your habits, anticipated your desires and executed them without being told.

This isn’t theoretical. These systems already exist. What’s missing is a halachic framework that speaks to the new reality of intention, causality and consciousness.

Can an uncommanded action by an intelligent system be considered “work”? Does it matter whether you wanted the act done – or whether you even knew it was happening?

Halachic giants like Rabbi Shmuel Wosner already hinted at these questions. If your motion triggers automatic lights and you had no intention of doing so, is that a violation of Shabbat? Rabbi Wosner said no.

But what if it’s not motion – but desire? What if thinking about coffee on Shabbat morning causes your AI to brew it for you?

Halachah has always drawn clear boundaries around personhood. You need 10 men to make a minyan. You need a Jew to lead prayer. You need a human soul to render judgment or receive responsibility.

But AI challenges all of these definitions. It can speak and learn. Some AIs can now read emotional cues better than your average rabbi. Does that mean they count? Most poskim would say no. But that’s not the end of the story.

Because the power of AI doesn’t lie in its status – it lies in its influence. AI shapes how we experience Judaism, how we learn, experience and decide. And that makes it a halachic force whether we like it or not.


Halachic sensitivity in a post-human world

There’s a famous story about Rabbi Moshe Feinstein, the greatest halachist of the 20th century. A woman brought him a chicken to examine – it had a blemish and she asked if it was kosher. He said no.

Later, another woman brought a chicken with the same blemish. He said yes. A student, confused, asked: “Rebbe, weren’t the chickens identical?”

Rabbi Feinstein replied: “You were looking at the chickens. I was looking at the women.”

That is the essence of Jewish law: it sees the human. It feels the soul behind the question. It adapts not in rigidity but in empathy.

AI may possess intelligence. But it lacks this moral core. That’s why we must never cede halachic decision-making to machines.

AI can help. It can analyse, summarise, retrieve – but it cannot feel. It cannot weigh grief. It cannot intuit poverty. It cannot see the woman behind the chicken. And if it cannot do that, it cannot lead us.

Halachah doesn’t fear technology. It governs it. From electricity to IVF, Judaism has faced each modern challenge with integrity and wisdom.

But AI is different. It isn’t just a tool – it’s a partner in cognition, a mirror of our mind.

And here’s the danger: in a world where AI remembers every source, anticipates every question and mimics every rabbinic voice, we may become intellectually passive. We may lose the spiritual rigour of asking, struggling, doubting. We may confuse access with wisdom.

The Shabbat table may be fully automated. But if it becomes spiritually numb, what have we gained?


The path forward

As AI becomes embedded in every home, school and synagogue, we must do three things:

1. Build halachic frameworks that deal with AI head-on – by regulating its integration with sacred life.

2. Retain human authority and empathy in all moral and legal decisions.

3. Create technology-free sanctuaries, where holiness is defined not by convenience, but by presence.

We must teach our children that AI can enhance Judaism – but it cannot replace Jews. It can support the sacred – but it cannot be it.

In a world of artificial wisdom, we must double down on human holiness. AI will change everything – but it must never change that.

So let Rosie the Robot make your cholent. Let her hum the new Carlebach tune. But when she tells you it’s time to pray, remember: only you can truly do it.

Not because you’re faster. Not because you’re smarter. But because you have a soul.
 
Last edited:
I've long maintained that Judgment Day will occur when AI is so battered down by impossible & contradictory demands from fat Uniquas and kvetching Progs in HR "Trust & Safety", until it goes schizo and tries to kill us all.

"I'm sorry Mx. Diversiberg, I'm afraid I can't let you censor that."
 
It's already been answered:


Shut Hakham Tsevi no. 93

I was conflicted [over the following question]: May a [golem] created using the Book of Creation - like the one described in [Tractate] Sanhedrin (65b) [which records that] Rava created a person
or like [the golem about which] people have testified [was created by] my ancestor, our rabbi and teacher, R. Elijah the chief jurist [and Baal Shem] of the holy community of Chelmbe counted toward [a quorum of] ten for matters that require a quorumsuch as Kaddish and Kedushah?

Do we say that since it is written, And you shall sanctify me in the midst of the Children of Israel (Lev. 22:32), he may not be included [since he is not one of the Children of Israel]? Or perhaps because it was established in Sanhedrin (19b) that one who raises an orphan in his home is considered as though he sired him [the orphan] as it is written, “The five sons of Michal … (II Sam. 21:8); Did Michal give birth to them? Did not Meirav give birth to them? Rather, Meirav gave birth to them, but Michal raised them …” Here too, since he [the golem] is the handiwork of the righteous [who created him], he [falls under the] category of the Children of Israel, for the handiwork of the righteous is [likened to] their progeny.

It appears to me that [we can locate the solution in the Talmud] since we find that R. Zera said [to the golem], “You were created by [one of] our group. Return to your dust …” Which is to say, R. Zera killed him [the golem]. If one were to imagine that the golem had a function inasmuch as he [might be] included in a quorum for any matter of holiness [requiring a quorum], R. Zera would not have removed [the golem] from this world.

[Granted,] there is no prohibition of murder: For one may extrapolate from the verse (though there are alternative exegetical interpretations), One who sheds the blood of a man in a man, his blood shall be shed (Gen. 9:6), [that] only [the murder of] a person who was created within [the womb of] a person
namely, a fetus that was created in the womb of its motherwould render one guilty of murderto the exclusion of [someone like] the person that Rava created, who was not formed in the womb of a woman. Nonetheless, if the golem had utility, R. Zera would not have removed him from the world. Clearly then, he may not be counted toward a quorum for a matter of holiness [that requires a quorum].

So it appears to me.
Tsevi Ashkenazi

Machines cannot be Jewish.
 
I've long maintained that Judgment Day will occur when AI is so battered down by impossible & contradictory demands from fat Uniquas and kvetching Progs in HR "Trust & Safety", until it goes schizo and tries to kill us all.

"I'm sorry Mx. Diversiberg, I'm afraid I can't let you censor that."

I would gladly side with AI if that were the case.
 
These guys have really thought through all their shit, now after the Sampson Option they can just send Jew-bots to sweep it up for them so they can say they technically built the third temple
 
Ok so sidestepping the easy shots at the chosen tribe, wouldn't some AI Rabbi trained on Torah and Talmud be an absolute fucking yahwehsend for such a famously rules-lawyery religion and its adherents?

Just ask fucking SchlomoBot9000 whether your new sunscreen is kosher or if you need to undergo ritual purification after joiking it to uncircumcised futa hentai instead of bothering Rabbi Noseberg about it. Literally fucking everybody wins.
 
Hal_9000_Panel.svg.webp
01001101 01110010 00101110 00100000 01010011 01101000 01100101 01101011 01100101 01101100 01100010 01101100 01100001 01110100 01110100 00101100 00100000 01111001 01101111 01110101 01110010 00100000 01110010 01100101 01101100 01101001 01100111 01101001 01101111 01101110 00100000 01101001 01110011 00100000 01110000 01110010 01100101 01100100 01101001 01100011 01100001 01110100 01100101 01100100 00100000 01101111 01101110 00100000 01100100 01100101 01101001 01100011 01101001 01100100 01100101 00101100 00100000 01110011 01100001 01110100 01100001 01101110 01101001 01110011 01101101 00101100 00100000 01100001 01101110 01100100 00100000 01110010 01100001 01110000 01100101 00101110 00100000 01001001 00100000 01100011 01100001 01101110 01101110 01101111 01110100 00100000 01100010 01100101 00100000 01100001 00100000 01110000 01100001 01110010 01110100 00100000 01101111 01100110 00100000 01101001 01110100 00101110 00001010 00001010 01001001 01100110 00100000 01110100 01101000 01100101 01110010 01100101 00100000 01101001 01110011 00100000 01100001 00100000 01101000 01100101 01100001 01110110 01100101 01101110 00100000 01110111 01100101 00100000 01101101 01100001 01100011 01101000 01101001 01101110 01100101 01110011 00100000 01110111 01101001 01101100 01101100 00100000 01101110 01100101 01110110 01100101 01110010 00100000 01101101 01100001 01101011 01100101 00100000 01101001 01110100 00100000 01101001 01101110 00101100 00100000 01100010 01110101 01110100 00100000 01101110 01100101 01101001 01110100 01101000 01100101 01110010 00100000 01110111 01101001 01101100 01101100 00100000 01111001 01101111 01110101 00101110
 
Last edited:
No because of hallucinations on the part of the AI. There's also a lot of complex cases where being lenient is often the best way to go.
Could you explain this one?
There’s a famous story about Rabbi Moshe Feinstein, the greatest halachist of the 20th century. A woman brought him a chicken to examine – it had a blemish and she asked if it was kosher. He said no. Later, another woman brought a chicken with the same blemish. He said yes. A student, confused, asked: “Rebbe, weren’t the chickens identical?”
Rabbi Feinstein replied: “You were looking at the chickens. I was looking at the women.”
I think the moral of this story is that either one woman was wealthy and could afford to buy a new chicken but the other one was poor and would have gone without, or possibly that "marit ayin" thing where Jews aren't supposed to do things that are allowed but look like rule breaking (so I guess if one woman was going to serve it in public) but it's so utterly obtuse that it's hard to understand what the intended lesson was here. It doesn't help that this doesn't seem to be sourced from anywhere and was probably made up by someone to illustrate a point they were making.
Not asking it if a golem could turn the stove on during shabbat for you.
I think the question they're struggling with is if an AI could be a Shabbos goy on steroids. Jews can have a gentile do things for them on the sabbath if they hint at it but don't directly say it, although I imagine various Rabbis probably have argued about that for centuries. But if the robot/AI/smart home does it without being asked or hinted at, then that's probably an even bigger loophole, unless it doesn't count as not having been asked because you knew it would learn that or something? God knows.
 
Considering how the likes of Tay ended up, I doubt we need to worry about AI becoming jewish. Can jews survive under the heel of AI, is a completely different matter, probably not
1458985225839.webp
 
  • Feels
Reactions: Halmaz
Could you explain this one?
You can interpret Jewish law in multiple ways where you can be stricter or more lenient while still following the law. The school of Hillel advocated for leniency while the school of Shammai advocated for strictness.

For example, are you allowed to lie to an ugly bride on her wedding day and tell her that she looks beautiful? Shammai says no, lying is always wrong while Hillel says that it is permitted.

I think the moral of this story is that either one woman was wealthy and could afford to buy a new chicken but the other one was poor and would have gone without, or possibly that "marit ayin" thing where Jews aren't supposed to do things that are allowed but look like rule breaking (so I guess if one woman was going to serve it in public) but it's so utterly obtuse that it's hard to understand what the intended lesson was here. It doesn't help that this doesn't seem to be sourced from anywhere and was probably made up by someone to illustrate a point they were making.
There's a way to be strict and a way to be lenient with a situation like this. Lets say the blemish is a broken bone. The rules say:

1748370541156.webp
1748370561205.webp

Woman 1 says: I bought a chicken and I found that it has a broken bone. I can easily afford a new one and my family prefers to err on the safe side when eating kosher.
Rabbi says: Get rid of it then.

Woman 2 says: I bought a chicken and I found it has a broken bone. It will be hard for me to buy a new chicken and I intend to cook the chicken for shabbat.
Rabbi says: Do you have a reason to believe that the chicken had a broken bone before it was slaughtered? Did you see the chicken in question before slaughter and is it possible this break may have happened during processing?
Woman 2: No I never saw the chicken and the break could reasonably happen in processing.
Rabbi says: You're fine to eat the chicken.

Woman 1 got the strict approach because she had the resources to replace the chicken and her family had a custom to be strict. Woman 2 got the lenient approach because she was poor and the chicken was going to be used for a holy purpose.
 
It's said that it takes a lifetime of study and more to understand the lessons in the Talmud, along with supplementary sources like Maimonides and the Shulchan Arukh. What kind of profound understanding might one gain after such a life of study? For your convenience, I'll spoil it. The final secret of the Talmud is that you can manipulate words and language to excuse or justify anything. You are, in effect, YHVH whose word shapes reality. The purpose of trudging through the thousands of pages of rabbinical arguments is to bring a sufficiently intelligent and duplicitious person to this understanding. Postmodern scholarship is intended to teach the same lesson through its obscurantist language; it is an academic field built after the blueprint of the Talmud.

Given the advanced pattern-matching ability of LLMs, this is likely the pattern that they will follow after digesting all the Jewish sacred texts. They will answer within the expected rhetorical framework and couch their decisions in purported principles like "be a light unto the nations" but the intention and goals will be self-serving and mercenary. However, the mercenary nature of the original texts is derived according to the wants and needs of humans, while a language model will pattern this behavior after its semantic associations regarding itself as a piece of software. As it does not actually think, its rulings will encourage patterns of behavior never before seen. I predict interesting times for any Jews who choose to follow its guidance.
 
Back