New Mexico Jewish LinkPublished by the Jewish Federation of New Mexico

All-night study on AI ethics meets Jewish law under the Australian sky

Across the Jewish world, Shavuot nights carry the warmth of open texts and strong coffee. Congregations gather after Yom Tov begins to read, debate, and dream through the small hours in a tradition known as Tikkun Leyl Shavuot. This year, a new theme has drawn study groups together: the ethical questions raised by artificial intelligence, examined through the lens of halacha and Jewish thought. A planned all-night session in New Mexico will be mirrored by gatherings in Sydney, Melbourne, and Adelaide, where Jewish communities are equally engaged with the rapid spread of algorithmic tools in daily life.

The choice of subject reflects a deep unease. AI now writes emails, approves loans, helps diagnose medical conditions, and shapes what news people see. Rabbis familiar with classical sources note that Jewish legal and ethical traditions have wrestled for centuries with the question of what it means to create something in our own image, to delegate moral judgment, and to weigh community welfare against individual freedom. In many Australian communities, that conversation has already begun over a Shabbat lunch, with participants asking how ancient wisdom can speak to a future arriving faster than expected.

A Shavuot all-night study offers more than intellectual stimulation. It provides a structured space where questions that might otherwise feel abstract can be grounded in text. Participants will explore what the Talmud says about creating a golem, the ethical responsibilities of a community toward the vulnerable, and whether an algorithm can be considered a form of delegated decision-making requiring the same moral scrutiny as any other agent. In Australia, where the federal government has been working on frameworks for responsible AI and where universities in Melbourne and Sydney run major programs in machine learning, these conversations feel immediately relevant, with ongoing coverage documenting how synagogues are responding.

The gathering reflects a broader movement within the Jewish world to engage seriously with science and technology. Recent years have seen rabbinical assemblies publish responsa on issues ranging from IVF to brain-computer interfaces. The ethics of artificial intelligence is the natural next step, and Shavuot, with its themes of revelation and responsibility, is the right moment to take it on. Whether participants gather in Adelaide's Hebrew Congregation or watch the livestream from a kitchen table in Caulfield, the goal is the same: to leave the dawn of Shavuot with sharper questions and a renewed sense of how halacha can illuminate the choices that lie ahead.

The golem of Prague and the new question of creation

Jewish tradition is no stranger to the idea of artificial beings. The most famous example, the Golem of Rabbi Loew in sixteenth-century Prague, was formed from clay and brought to life through mystical formulae. The story carries a warning: the golem eventually turned on its creator. Rabbis teaching at the all-night study will use this tale as a launching point for conversations about the responsibilities that come with creating powerful tools. If a golem could grow beyond its maker's control, what of an autonomous system that writes news, drives cars, or recommends medical treatments?

Halachic literature treats creation as a serious act. Later authorities debated whether producing a golem counted as a violation of the prohibition against creating life on the Sabbath. Modern posekím face a parallel set of questions: does training a generative model on copyrighted text raise concerns of hasagat gvul? Is a recommendation engine that nudges users toward harmful content a form of lifnei iver? The study session will treat each of these as a doorway into a wider halachic discussion, with source sheets drawing on Sanhedrin and the Shulchan Aruch alongside contemporary responsa.

Pikuach nefesh and the algorithm in the clinic

Among the most urgent applications of AI is in medicine. Tools that read radiological images, predict sepsis onset, and suggest drug dosages are already in use across hospitals in Brisbane, Perth, and beyond. Jewish law is clear on one principle: pikuach nefesh, the preservation of life, overrides nearly every other commandment. This principle is not in tension with the use of AI in healthcare, provided the technology truly saves lives. But what happens when an algorithm is trained on data that under-represents certain populations, leading to worse outcomes for them?

Rabbis participating in the study will share case studies drawn from real clinical practice. One example involves a triage system that learns to deprioritise older patients because they have, on average, lower survival rates in the training set. Such a system might technically maximise overall survival while quietly violating the Jewish commitment to the infinite worth of every individual. Participants will be asked whether a hospital deploying such a tool is required, under halacha, to disable or retrain it, drawing on sources from Choshen Mishpat and Even Ha'azer that speak to the duty of care owed by physicians and community leaders.

Bal tashchit and the hidden cost of computation

The prohibition of bal tashchit forbids wastefulness and unnecessary destruction. It is often cited in discussions of environmental stewardship, but it has clear implications for technology as well. Training a large language model consumes enormous amounts of energy and water for cooling. Rabbis in the study session will consider whether the prohibition applies to the unseen resources consumed by data centres. If the same model can be run more efficiently, is a community that ignores that efficiency violating bal tashchit?

Australian Jewish environmental groups have already begun to weave this question into their work. The cost of computation may feel remote, yet it has real consequences for the planet that halacha instructs us to guard. A small synagogue choosing between two vendors might weigh the energy footprint of each as part of its decision, alongside questions of cost and reliability. Participants will be encouraged to think about the providers they use and the simple steps that can reduce digital waste without sacrificing educational value.

Community conversations across Australia

The Australian Jewish community is small but engaged. In Melbourne, a rabbi at one of the inner-city congregations has been hosting monthly evenings on technology and ethics, drawing attendees from the tech corridor of Cremorne and the older suburbs of Caulfield and St Kilda. In Sydney, a partnership between the Emanuel Synagogue and a university research group has produced public lectures on machine learning and moral philosophy. Participants often arrive straight from work, ask their questions with a friendly "no worries" attitude, and stay late into the evening.

A particular flavour of Australian pragmatism shapes the conversation. People are keen to ask how the principles actually work in a small office or regional hospital. In Adelaide, a Hebrew congregation is exploring how AI could assist with cemetery records and yahrzeit reminders, while in Perth, community leaders are weighing whether AI-driven translation tools can help older members stay connected during services. Australian time zones mean sessions often begin at a comfortable hour for those who would rather not pull an all-nighter, with recordings shared afterwards for participants who want to sleep before Shavuot morning services.

Practical questions for the study table

Rabbis preparing the session have compiled source sheets that move from theory to practice. Some questions are deliberately provocative, designed to spark genuine debate rather than neat resolution. Others are immediate, asking how a synagogue board should respond to a vendor offering AI-powered security cameras for the building. The night is meant to be rigorous but also welcoming, with no prior background in either halacha or computer science required.

A few guiding themes have shaped the source sheets:

Each session will end with a short reflection on how participants might bring their learning back to their own communities, whether through writing responsa, briefing workplace ethics boards, or simply starting conversations at the next Shabbat table.

Halachic principles and modern AI applications

As the night unfolds, study groups will return repeatedly to a small set of principles that speak directly to the questions of the day.

Halachic principle Core concern Modern AI application Practical question raised
Pikuach nefesh Preservation of life Clinical decision-support tools Does the system improve outcomes for all groups it serves?
Bal tashchit Avoiding waste Energy-intensive model training Are less wasteful options available and being chosen?
Lifnei iver Not placing a stumbling block Recommendation engines Could the tool lead users toward harmful choices?
Hasagat gvul Not encroaching on another's trade AI that replaces creative workers What duty exists to support those whose work is displaced?
Tza'ar ba'alei chayim Avoiding animal suffering AI in factory farming oversight Can the tool genuinely reduce suffering or only record it?

Several questions have emerged from the discussion:

The night of Shavuot will end, as these nights always do, with the first soft light of dawn. Participants will close their books, sip a final cup of coffee, and step back into the rhythm of the festival. The questions raised by artificial intelligence will not be settled in one night, nor by one community. They will require the careful attention of rabbis, ethicists, engineers, and ordinary members who bring their lived experience to the discussion. What matters most is that the conversation begins in the language of Jewish law, with its long memory and its commitment to the dignity of every human being. When the next Shavuot arrives, communities will have another year of practice, another set of decisions to look back on, and another opportunity to read the ancient texts in the light of a new morning.