AI Meets Jewish Ancestry: A Workshop for Australian Researchers

Across living rooms in Caulfield, Bondi, and St Kilda, many Australian Jews are discovering that their family's pre-war story has quietly faded from living memory, leaving only a few faded photographs and a name or two on a Ketubah. A new workshop concept aims to change that, inviting community members to learn how modern artificial intelligence can help piece together fragmented family trees, translate Yiddish letters, and read handwritten entries in old synagogue ledgers. The idea, developed as part of a broader initiative that also explores supporting Jewish artists in residence at community centers, is pitched at anyone who has ever wondered whether their great-grandparents arrived in Melbourne through Sydney's docks, or which shtetl in Galicia their family left behind.

The workshop is designed to feel less like a lecture and more like a communal research session, the kind of gathering that might once have happened over Sunday lunch at a Deli in Elsternwick. Participants are guided through the basics of AI tools, given practical time with sample datasets, and encouraged to bring their own family documents to work on in real time. For Jewish Australians in particular, where records often span continents, languages, and political borders, this kind of guided hands-on support could be the difference between a brick wall and a breakthrough.

Why Jewish Genealogy Resonates in Australia

Australia is home to one of the world's most established Jewish communities outside Israel and the United States, with roots that stretch back to the first shiploads of free settlers and gold rush immigrants in the 1840s. By the time the great wave of post-war migration reshaped suburbs like Sydney's eastern beaches and Melbourne's inner south-east, Jewish life here had already developed its own distinctive character, anchored by institutions such as the Sydney Jewish Museum and the Jewish Holocaust Centre in Elsternwick. Yet for all the institutional memory held by these centres, individual family stories are often startlingly thin, particularly among descendants of Holocaust survivors who arrived with very little documentation of their own.

This thinness of record is exactly what motivates genealogists to keep digging. A person might know that their grandfather arrived in Sydney from Bucharest in 1949, but have no idea what village he came from, what his original surname might have been, or whether any cousins survived the war in France. Each of these missing details represents a potential doorway into a richer understanding of self. The Australian Jewish Genealogical Society has been quietly filling in such doorways for decades, publishing journals and running beginner courses that connect researchers with records held in places as varied as the State Library of Victoria, the National Archives of Australia in Canberra, and synagogue archives in Perth.

The new workshop sits within this tradition but adds a contemporary layer. It treats AI not as a replacement for careful archival work, but as a fast-working research assistant that can read handwriting, transliterate Hebrew tombstones, summarise Polish-language civil records, and suggest likely connections across enormous databases. For a community that has always placed a high value on memory, education, and the continuity of story, the timing feels right.

What AI Brings to Family Tree Research

For decades, family history research has relied on patient reading of microfilm, long letters to overseas record offices, and the slow accumulation of paper certificates. AI does not so much replace that patient work as compress and accelerate parts of it. Optical character recognition models can now read faded entries from Polish civil registries, often producing clean transcriptions from images that would previously have required a fluent speaker. Large language models can summarise long, rambling Holocaust survivor testimonies in Yiddish or Polish, helping researchers extract names, places, and dates from sources that were effectively inaccessible a decade ago.

The workshop walks participants through how these tools actually behave in practice. Rather than describing capabilities in the abstract, facilitators demonstrate tasks such as extracting a full list of names from a scanned Knesseth Israel burial register from Warsaw, or generating a translation of a 1920s Yiddish wedding invitation found in a Sydney attic. Participants learn when a model can be trusted, when its output needs verification against a primary source, and how to prompt it in ways that yield useful results. The emphasis throughout is on humility: AI is brilliant at certain narrow tasks and unreliable at others, and knowing the difference is the key skill.

For Jewish researchers in particular, one of the most promising developments is the rise of specialised tools trained on Hebrew-script handwriting and Yiddish transcription. These models, several of which are open-source, can handle the peculiarities of rabbinical script that often defeated earlier generations of genealogists. Combined with general-purpose AI services, they are opening up entire archives that have effectively been locked away in linguistic glass cases.

A Practical Workshop Format for Local Participants

The format under discussion borrows from the way Australian adult education has traditionally worked, with small in-person cohorts meeting over several weeks rather than a single overwhelming session. In Melbourne, sessions would likely be hosted at a community centre in Caulfield or in the rooms of a long-established congregation, where parking is easy and the venue is already familiar. In Sydney, similar sessions might run out of the Jewish House complex in Bondi or through the University of Sydney's adult learning program. Online participation from regional areas, including researchers in Brisbane, Adelaide, and even isolated mining communities in Western Australia, is also part of the plan.

Each session is structured around a small group, no more than twelve participants, so that everyone gets individual attention on their own family puzzle. Facilitators circulate, helping people craft prompts, interpret model outputs, and cross-reference findings with traditional databases such as Ancestry, MyHeritage, and the specialised holdings of the Australian Jewish Genealogical Society. A dedicated volunteer librarian is on hand to suggest archive boxes, microfilm reels, and books that might fill in the gaps the AI cannot yet address. Participants leave each week with at least one new piece of verified information about their family, however small.

The workshop is also being designed with intergenerational pairs in mind. Where possible, older relatives who hold the oral history are invited to attend alongside younger members who bring the technical skills. This pairing produces rich conversations, and it is not unusual for a grandparent to remember a half-forgotten name the moment a grandchild pulls up an AI-generated translation of an old document.

Working With Australian Archives and Records

The local flavour of the workshop matters enormously, because Australian Jewish records have their own quirks. Naturalisation papers from the early twentieth century are held in Canberra, but the corresponding shipping manifests live in state archives in Sydney and Melbourne. Synagogue minute books from places like the Melbourne Hebrew Congregation or the Great Synagogue in Sydney are often only partially digitised. Cemetery records for Rookwood, Springvale, and Willesden Jewish cemeteries exist in a patchwork of formats, with some maintained by descendants and others still held only in paper form.

The workshop dedicates an entire session to navigating this patchwork. Participants learn where the State Library of New South Wales keeps its collection of community histories, how to request files from the National Archives, and which genealogical societies hold duplicates of records held overseas. They also learn how to upload their AI-assisted findings to Australian platforms such as Trove, ensuring that new discoveries feed back into the wider community of researchers rather than disappearing into a personal hard drive.

Equally important is the lesson of how to read Australian records in their proper context. A naturalisation certificate from 1907 might list a place of birth that no longer exists on modern maps, having been renamed during twentieth-century border changes. AI tools can help suggest modern equivalents, but only when researchers prompt them carefully and verify the results against an atlas. The workshop treats these checks as a habit to be built rather than an afterthought.

Ethics, Accuracy, and Family Sensitivities

No workshop on AI and genealogy would be complete without a serious conversation about ethics. Family history is rarely purely academic. A discovery of a previously unknown relative, a wartime tragedy, or a long-buried secret can send shockwaves through a living family. Facilitators therefore spend meaningful time on how to handle sensitive findings, when to pause before sharing results with elderly relatives, and how to record sources carefully so that no claim is ever presented as more certain than the underlying evidence supports.

There is also a community-level dimension to consider. Many Australian Jewish families include members who are the sole survivors of entire branches, and a piece of new information uncovered by one researcher might be the missing piece another family has been searching for decades. The workshop encourages participants to share findings through established community channels, including the Australian Jewish Genealogical Society's journal and online forums, and to treat living relatives, including those discovered through research, with respect and care.

AI tools introduce their own ethical wrinkles. Models trained on large swathes of internet text can occasionally fabricate plausible-sounding details, a phenomenon researchers call hallucination. The workshop devotes part of one session to building a habit of verification, teaching participants to treat AI output as a starting hypothesis rather than a final answer. Every name, date, and place is checked against at least one independent source before it enters the family tree.

Turning Findings Into Living Family Stories

The ultimate aim of all this careful work is not just an expanded family tree but a strengthened sense of identity and belonging. Workshop organisers encourage participants to write up their findings, however modest, into short family narratives that can be shared at Sedarim, b'nai mitzvah gatherings, and family reunions. Several Australian rabbis have already expressed interest in incorporating these stories into Yom HaShoah commemorations, particularly in Melbourne, where the annual march from the State Library to the Melbourne Holocaust Museum draws thousands.

There are also practical ways the workshop feeds back into community life. Some participants go on to volunteer at local Jewish historical societies, helping to transcribe and translate records that have languished unread for decades. Others become informal mentors for newcomers, sharing the prompts and tools they have found most useful. The result, over time, is a community that becomes collectively better at the work of remembering, with younger generations equipped to carry the tradition forward using tools their grandparents could never have imagined.

If you would like to see this idea move from concept to program, you can support it through The Next Big Jewish Idea community campaign. Vote for the proposal, share it with friends in your community, or reach out to the organising team to offer your skills as a facilitator, translator, or archival volunteer.