Ancestry has been applying machine learning to genealogy for more than a decade — the Hint system that surfaces matching records in your tree has used algorithmic matching since its earliest versions. But the past two years have seen a qualitative shift, as Ancestry has rolled out a suite of features explicitly branded as AI-powered and directed at a general audience rather than the research-oriented core. The result is a mixed picture: some of these tools represent genuine advances in what a genealogy platform can do, while others trade on the AI label to offer something considerably more modest. Separating the two is worth doing carefully.

AI Stories: Genuinely Useful Within Its Limits

The flagship AI feature, and the one Ancestry has promoted most heavily, is AI Stories. AI Stories is currently available on over 9.8 billion records, turning static historical documents into engaging narratives that bring context and meaning to the facts. The feature is available in six languages — English, Swedish, German, French, Spanish, Dutch, and Italian.

What this actually does: when you view a historical record — a census entry, a ship manifest, a draft registration card — an AI-generated paragraph appears alongside it, placing that document in its historical context. A 1918 draft card for a Dillman ancestor doesn’t just show his name and address; the AI Story paragraph explains what was happening in the war at that moment, what the draft registration process involved, and what daily life looked like for men of his age in that period.

For serious researchers, this sounds like noise. For family members who aren’t genealogists but are curious about ancestors — the people you’re ultimately building a family history for — it is genuinely valuable. It transforms a bureaucratic form into a readable moment in a person’s life. Used for what it actually is (context and engagement, not genealogical evidence), it earns its place.

What it is not: AI Stories does not synthesize information across records, does not make genealogical connections, and does not draw conclusions about your specific ancestor. It provides generic historical context, not individualized research. Its citations are to general historical knowledge, not to the specific record being displayed.

Document Transcription: The Most Useful AI Feature of All

Document Transcription leverages AI to read journals, letters, postcards, and other handwritten family heirlooms, accurately transcribing even the most challenging-to-read handwriting and making them searchable, unlocking insights that might otherwise remain hidden.

This is, without qualification, the most practically useful AI tool Ancestry has released. The transcription of handwritten documents has been one of the most significant bottlenecks in genealogical research for as long as the field has existed. Nineteenth-century German Kurrent script, Irish parish registers in faded ink, Civil War pension applications in a clerk’s cramped hand — all of these have traditionally required significant paleographic expertise to read. AI transcription doesn’t replace that expertise, but it now makes a first pass that is accurate enough to be genuinely useful.

For Dillman researchers specifically, the ability to upload photographs of handwritten documents — including German-language church records, land grants, and correspondence — and receive a searchable transcription reduces a significant barrier to research. The technology is not perfect, particularly with non-English languages and unusual handwriting, but it has improved substantially year over year.

Face Match: Interesting but Handle With Care

Face Match uses AI to scan Ancestry’s database of images and identify photographs that may depict the same person. The stated goal is to help researchers “put a face to a name” when a family photograph can be matched to other images in the database.

AI photo restoration of portraits still feels problematic, with many people satisfied with an artificially generated “average” set of features taking the place of the unknown features of an ancestor on a darkened old print. Without meaning to seem over-dramatic, the historical record has already likely been polluted by these images, since we can’t trust all users to watermark or label them appropriately.

Face Match operates differently from photo restoration — it identifies potential matches between existing photographs rather than generating new imagery. The concern is similar, however: facial recognition applied to degraded nineteenth-century photographs is not reliable enough to use as genealogical evidence, and there is a real risk that researchers will treat a Face Match result as confirmation of identity when it is at best a hypothesis. Use it as a search tool to surface photographs that might be the same person, then verify through documentary means. Do not use it as evidence.

AncestryPreserve, Full Text Search, AI Ideas, and Networks

Ancestry rolled out four additional features as part of its ProTools suite: AncestryPreserve, Full Text Search (beta), AI Ideas, and Networks (beta).

AncestryPreserve allows users to upload and store family documents, photographs, and memorabilia directly in the platform — not as research sources attached to specific people, but as a preservation archive linked to the family tree. This is a practical response to a real problem: genealogical research materials are often scattered across personal devices and platforms, and centralized preservation in a reputable platform is genuinely valuable. The caveat: your documents are on Ancestry’s servers, which returns us to the data ownership concerns addressed in our earlier privacy article.

Full Text Search (beta) extends Ancestry’s search capability from record index fields (name, date, place) to the full text of digitized documents. This means that a name appearing in a body of text — an obituary mentioning a sibling’s name, a pension affidavit describing a neighbor — can now be found even when it wasn’t captured in the original indexing. For serious researchers, this is potentially significant: it surfaces connections that conventional indexed search completely misses.

AI Ideas analyzes your tree and suggests research directions — “consider looking for a death record for this individual,” “this person appears in a census that may contain additional family members.” For experienced researchers, this is nothing new. For beginners, it may provide useful direction. Its value scales inversely with research experience.

Networks (beta) is the most conceptually distinctive of the four. Where the rest of Ancestry’s toolset focuses on the documented family tree — blood and legal relationships — Networks is designed to map the broader community context that shaped your ancestors’ lives. It allows researchers to create and organize groups of connected individuals who are not necessarily related: neighbors from a census record, members of the same church congregation, passengers on the same ship, soldiers in the same regiment. This is what genealogists call “FAN Club” research — tracking Friends, Associates, and Neighbors alongside family members — and it is one of the most productive strategies for breaking through brick walls. Ancestors who can’t be found in isolation are often traceable through the communities they traveled with.

The tool is fully customizable: researchers can define their own network types and connection categories. One significant limitation noted by early users is that records cannot be cited at the network level — a gap that matters for anyone who wants their community research to be as rigorously documented as their tree. That limitation is likely to be addressed in future updates as the beta matures.

What AI Cannot Do for Genealogical Research

The honest summary of Ancestry’s AI suite is that it excels at making research more accessible to non-specialists and at automating tasks currently handled by search indexers and translators. Ancestry’s ThruLines and Hints features use AI to surface potentially relevant records from their 40+ billion record collection and connect your tree to those of other users. The AI is not infallible — it suggests connections, and you must verify.

This last point is the crucial one. No AI feature on any platform currently produces genealogical conclusions that can be trusted without verification. The records being surfaced are real; the connections being suggested may or may not be. The standard of proof required for genealogical conclusions — source citation, correlation of evidence, resolution of conflicts, reasonably exhaustive search — remains a human responsibility. AI accelerates the search; it does not do the research.

For Dillman Researchers

The AI features most relevant to Dillman-specific research are Document Transcription (for German-language records and handwritten materials) and Full Text Search (for finding name occurrences in unindexed text). Networks has particular appeal for Dillman research given the cluster-migration patterns the DFA has documented: when a group of Dillman-variant families left the same Alsatian or Hessian village within a decade of each other and settled in the same American county, building a Network around that community reveals connections that a standard family tree cannot capture. AI Stories provides useful historical context for American records but is limited for German-origin research, where the contextual framework is less well-developed. Face Match should be treated as a hypothesis generator, not evidence.

Sources

Ancestry. “Looking Back on 2025: A Year of Growth, Discovery, and Innovation at Ancestry.” Ancestry Corporate Blog. Accessed July 2026. ancestry.com.

Ancestry. “Ancestry Brings Family History to Life with New AI-Powered Stories.” Ancestry Blog. December 15, 2025. ancestry.com.

Perl, Jonny. “Genealogy and DNA: A Review of 2025.” DNA Painter Blog. January 30, 2026. blog.dnapainter.com.

Family History Foundation. “New Ancestry Features in 2026: How to Effectively Use AncestryPreserve, Full Text Search & AI Tools.” March 12, 2026. familyhistoryfoundation.com.

Genealogy TV. “Ancestry New Networks: A Powerful New Research Tool.” March 6, 2025. genealogytv.org.

Family History Fanatics. “Does Ancestry’s New Networks Tool Actually Help? Here’s My Honest Review.” May 1, 2025. familyhistoryfanatics.com.

Incarn. “11 Best AI Tools for Genealogy Research in 2026.” June 1, 2026. incarn.co.

Note: This post was researched and written with the assistance of artificial intelligence (AI). All content has been reviewed for accuracy as of September 2026. All errors are the responsibility of the author.