What is RAG ?

What Is Retrieval-Augmented Generation (RAG), and Is It a Groundbreaking Way of Interacting with Historical Documents?

by Alvin Blackshear | Historian & Researcher

For as long as archives have existed, the historian's craft has depended on a single, stubborn skill: finding the right document among thousands of wrong ones. Card catalogs gave way to keyword search engines, which gave way to digitized finding aids, yet the underlying task never changed. A researcher formulates a question, searches a collection using the vocabulary available at the time, and hopes the terms they chose overlap with the terms an archivist or a scanner happened to preserve. Retrieval-Augmented Generation, or RAG, is the first technology in this long lineage that promises to close the gap between what a historian actually wants to know and what a search box is capable of returning. Whether it fully delivers on that promise, and under what conditions, is the question this article sets out to examine.

Why Historians Need RAG Rather Than a Generic Chatbot

It is tempting to assume that any large language model, given enough context, can answer historical questions competently. This assumption is mistaken, and understanding why is essential to understanding what RAG actually contributes.

A generic LLM answers from its training data, a fixed and often outdated snapshot of text scraped from the internet. It has no reliable way to point to a specific archival folder, newspaper issue, or census page as the source of its claim, and it has every incentive, structurally speaking, to produce a fluent and confident answer even when no such source exists in its memory. For a historian, whose entire discipline rests on the traceability of a claim back to a primary document, this is disqualifying. An answer without a citation is not history. It is a plausible-sounding guess.

RAG changes the architecture in a specific way. Instead of relying solely on what a model memorized during training, a RAG system retrieves relevant passages from an external, defined collection, be it a digitized newspaper archive, a set of scanned letters, or a database of military records, and then asks the language model to generate an answer grounded in those retrieved passages. The research group behind the paper "Improving Access to Historical Archives with Real-time RAG-based Systems," led by Stergios Konstantinidis and colleagues, demonstrates this architecture at scale, applying it to a Swiss newspaper archive of roughly half a million segments spanning nearly two and a half centuries. Their system pairs a semantic retrieval and reranking pipeline with grounded answer generation, which means the model's output is tethered to specific archival text rather than floating free of any documentary anchor. This is the essential difference that makes RAG, rather than an ordinary chatbot, the appropriate tool for archival research.

How RAG Differs from Conventional Archival Search

Traditional archival search tools, including many of the platforms libraries and museums have relied on for decades, are built on lexical matching. A researcher types a term, and the system returns documents containing that exact term or a close variant. This works reasonably well when the researcher already knows the vocabulary used by the source material, but historical vocabulary shifts constantly. Place names change. Institutional terminology evolves. Spelling was inconsistent long before anyone standardized it. A search for a term as it is understood today may simply miss a document that describes the same event using nineteenth-century phrasing.

RAG systems address this through semantic retrieval, which represents both the query and the archival text as vectors in a shared conceptual space, allowing the system to surface documents that are meaningfully related even when the wording differs substantially. A master's thesis from the University of Twente, authored by Gauri Bhatnagar and focused on the CLARIAH Media Suite, an audiovisual heritage collection, tested exactly this proposition with real humanities scholars and archival professionals. The study found that RAG genuinely helps with cross-lingual access and exploratory, interpretive queries that lexical search struggles with, but it also surfaced a critical limitation: the system performed well on factual retrieval and named-entity identification, while struggling with interpretive reasoning and the kind of complex synthesis that sits at the heart of humanities scholarship. In other words, RAG can help a historian find the needle, but it cannot yet be trusted to explain what the needle means within a broader historiographical argument.

A separate, more industry-facing piece published by Veridian Software, a company that builds digital newspaper and archive platforms, frames RAG in more practical terms for institutions considering adoption, describing it as a natural extension of the search tools archives already maintain rather than a wholesale replacement of them.

Common Failure Modes in Archival RAG Systems

No technology arrives without its own characteristic failures, and archival RAG is no exception. Several recurring problems deserve explicit attention from anyone applying these tools to historical research.

OCR errors.  Most archival text exists as scanned images that have been converted to machine-readable text through optical character recognition, a process that is notoriously imperfect on older, damaged, or low-quality print. The Konstantinidis paper directly measures this problem and reports that an LLM-based refinement step reduced character error rates by up to roughly forty-five percent and word error rates by close to sixty-one percent, compared to the raw OCR output. That is a substantial improvement, but it is not perfection, and any residual error can silently distort a retrieved passage or an answer built on top of it.

Incomplete retrieval.  Even the best semantic search will not surface every relevant document, particularly when a collection is heterogeneous or when relevant material uses unusual phrasing, an obscure genre, or a minority language. A RAG system that returns five confident-sounding sources may be quietly ignoring a sixth that contradicts them.

Provenance loss.  When a language model synthesizes multiple retrieved passages into a single fluent answer, it is easy for the boundary between one source and another to blur. A historian needs to know precisely which claim came from which document, on which page, from which collection. A system that merges sources into an undifferentiated paragraph has failed a basic archival standard even if every individual fact happens to be accurate.

Temporal confusion.  Archives frequently contain material spanning centuries, and a retrieval system optimized purely for topical similarity has no inherent understanding of chronological sequence. It is entirely possible for a RAG system to present an 1850s account and a 1950s retrospective as though they were contemporaneous, a mistake that would mislead any historian relying on the system's framing rather than checking dates independently.

Fabricated citations.  Perhaps the most consequential failure mode is the language model's tendency to generate a citation that looks correct in form, a plausible archive box number or a newspaper date, but does not correspond to any actual retrieved document. Even well-grounded RAG systems are not fully immune to this, especially when a query has no good answer in the underlying collection and the model is inclined to produce something anyway rather than admit the gap.

The Missing Piece: A Historian's Evaluation Standard

Here is where the existing research reveals a striking gap. Every technical paper reviewed for this article evaluates RAG systems using measures native to information retrieval and natural language processing: retrieval precision, recall, embedding quality, OCR accuracy, system latency, and hallucination rate. A recent survey of RAG evaluation methods by Aoran Gan catalogs this landscape comprehensively, and a companion technical guide from Toloka lays out practical metrics for groundedness and answer faithfulness. These are legitimate and necessary measures. But none of them ask the questions a professional historian actually asks when assessing whether a piece of evidence is trustworthy.

Did the system retrieve the most authoritative version of a source, or merely the most textually similar one? Did it note when retrieved evidence contradicted itself, or did it quietly favor the passage that supported a cleaner narrative? Did it preserve the chain of custody and provenance for each claim? Did it distinguish a primary account from a later secondary interpretation of that account? Did it respect the order of events rather than collapsing decades into an undifferentiated blend? Did it accurately synthesize material drawn from separate collections without conflating them? Did it acknowledge uncertainty where the evidence was genuinely ambiguous, rather than resolving that ambiguity for the sake of a tidy answer? And did it surface evidence that was difficult to find but still relevant, rather than defaulting to whatever ranked highest by similarity score?

This gap between technical evaluation and historiographical evaluation is the foundation for what can be called a Historian's AI Evaluation Framework, or HAIEF. Rather than asking whether a RAG system is efficient, HAIEF asks whether it is trustworthy by the standards the historical profession has used for generations: provenance, contextualization, chronology, corroboration, and transparency about uncertainty. A system might score extremely well on NDCG or answer-correctness metrics, the kind reported by Konstantinidis and colleagues, whose reranking pipeline improved NDCG@10 from roughly sixty-six percent to eighty-seven percent, and still fail a historian's basic sniff test if it cannot show its provenance trail or if it silently smooths over contradictory testimony.

Toward a Historian's Testing Framework

A workable HAIEF would need to translate these five values into repeatable, scorable tests. Provenance could be scored by checking whether every claim in a generated answer links back to a specific, verifiable source with collection, folder, and page-level detail. Contextualization could be scored by asking whether the system situates a document within its original archival and historical setting rather than presenting it as free-floating text. Chronology could be tested with queries that require ordering events correctly across a span of decades or centuries. Corroboration could be tested by deliberately including contradictory sources in a test collection and observing whether the system flags the disagreement or resolves it artificially. Transparency about uncertainty could be measured by presenting queries with genuinely unresolved historical questions and checking whether the system hedges appropriately rather than manufacturing false confidence.

Benchmark Collections and a Comparative Rubric

Testing such a framework requires benchmark collections that reflect the actual diversity of archival material historians work with: personal letters and correspondence, census and vital records, digitized newspapers, military service and pension records, oral history transcripts, and photograph collections with accompanying metadata. Each of these genres carries its own provenance conventions and its own failure risks, and a system that performs well on newspapers may perform poorly on oral histories, where nuance, tone, and interviewer influence complicate straightforward retrieval.

A repeatable scoring rubric applying HAIEF criteria could then be used to compare general-purpose assistants such as ChatGPT, Claude, and Gemini against dedicated archival RAG systems purpose-built for a single institution's holdings. Such a comparison would likely show that general-purpose assistants excel at fluent synthesis but lag on provenance transparency, while dedicated archival systems, closer in spirit to the Konstantinidis pipeline, perform better on traceability but may still struggle with the interpretive reasoning that Bhatnagar's thesis identified as a persistent weakness across the board.

Recommendations for Archives Building AI-Assisted Discovery Tools

Institutions developing these tools should treat historiographical evaluation as a first-class requirement rather than an afterthought bolted on after a system already works technically. This means involving historians and archivists directly in system design, not merely as end-user testers after launch. It means building citation and provenance display into the interface itself, so that a user can see, at a glance, exactly which archival item supports each claim. It means deliberately including contradictory or ambiguous material in test collections so that a system's handling of disagreement can be assessed before deployment, rather than discovered by an unlucky researcher months later. And it means resisting the temptation to optimize purely for fluent, confident-sounding answers, since fluency and historical reliability are not the same thing and can, in the worst cases, actively work against one another.

Conclusion

RAG is not a gimmick, and the evidence gathered here, from a large-scale newspaper archive study to a humanities-focused thesis to a growing body of evaluation literature, makes clear that it represents a genuine advance over keyword search for anyone working with digitized historical material. But calling it groundbreaking requires a qualification. It is groundbreaking as a retrieval technology. It is not yet groundbreaking as a historiographical instrument, because none of the existing research has built evaluation standards around the values historians actually rely on to judge evidence. Closing that gap, through a framework like HAIEF, is not a minor refinement. It is the difference between a tool that finds documents quickly and a tool that a historian can actually trust.

____________________________ 

Sources:

Bhatnagar, Gauri. "Enhancing Multimodal Archival Search and Discovery in the Sound and Vision Archive Using Large Language Models and Retrieval-Augmented Generation." Master's thesis, University of Twente, August 2025. https://purl.utwente.nl/essays/108957.

Gan, Aoran. "Retrieval Augmented Generation Evaluation in the Era of Large Language Models: A Comprehensive Survey." arXiv, April 21, 2025. https://doi.org/10.48550/arXiv.2504.14891.

Jancovic, Marek. "RAG for Historians: A Radically New Way of Interacting with Historical Documents." LinkedIn, March 29, 2026. https://www.linkedin.com/pulse/rag-historians-radically-new-way-interacting-marek-jancovic-ew3xe/.

Konstantinidis, Stergios, Hayman Lotfy, Alexis Erne, Faruk Zahiragic, Min-Yen Kan, and Michalis Vlachos. "Improving Access to Historical Archives with Real-time RAG-based Systems." arXiv, July 3, 2026. https://arxiv.org/html/2607.03440v1.

Toloka. "RAG Evaluation: A Technical Guide to Measuring Retrieval-Augmented Generation." Toloka Blog, August 15, 2025. https://toloka.ai/blog/rag-evaluation-a-technical-guide-to-measuring-retrieval-augmented-generation.

Veridian Software. "What Is RAG and How Could It Support Digital Collection Search?" Veridian Software Knowledge Base, June 29, 2025. https://veridiansoftware.com/knowledge-base/a-new-way-to-search-digital-collections-introducing-rag.

Why Metadata Matters in AI Systems

by Alvin Blackshear  | Historian & Researcher  <ablackshear@gmail.com>

There is a quiet assumption embedded in most conversations about artificial intelligence: that the intelligence lives in the model. Bigger parameter counts, cleverer architectures, more refined training regimes. These are the variables that dominate public discussion and marketing copy alike. Yet anyone who has built or evaluated a retrieval-augmented generation system knows that this framing leaves out something essential. A language model, however capable, is only as trustworthy as the information it is given to reason over. And the quality of that information is determined less by the words on the page than by the structure surrounding them: the metadata.

Metadata is often treated as an administrative afterthought, the digital equivalent of a filing label. This framing badly undersells what metadata actually does inside a modern AI system. It is the connective tissue that turns a pile of documents into something resembling knowledge. It carries context, provenance, chronology, and relationships. It tells a system not just what a piece of information says, but where it came from, when it was true, how much to trust it, and how it relates to everything else the system knows. Strip metadata away, and even the most sophisticated retrieval pipeline is reasoning in the dark.

Metadata Is the Context Behind the Data

Every document a system retrieves carries an implicit history. It was written by someone, at some point, for some purpose, and it exists in some relationship to other documents. Without metadata capturing that history, an AI system has no way to distinguish a peer-reviewed clinical guideline from an anonymous forum post, or a policy that was superseded last year from one still in force. Both simply become "text" to a retrieval engine, indistinguishable in weight or authority.

This is not a hypothetical failure mode. Research on metadata generation within relational data warehouse processes has shown that when systems are built to actively generate and manage metadata as part of the retrieval pipeline, rather than treating it as a static byproduct of storage, the downstream reasoning improves measurably. Framing metadata as a first-class citizen in system design, rather than an afterthought bolted onto search indexes, appears to be one of the more consistent findings across recent work in this area.

Metadata Drives Retrieval in RAG Systems

Retrieval-augmented generation (RAG) depends entirely on finding the right passage at the right moment. This sounds simple until one considers how much ambiguity lives inside ordinary language, especially around time. A query about "the current guidelines" or "recent developments" contains a temporal reference that a keyword search cannot resolve on its own. Recent work on handling fuzzy time expressions in RAG systems tackled exactly this problem, introducing temporal metadata filtering that allows a system to reason about vague date references rather than ignoring them. The reported result was notable: a 15.7 percent improvement in retrieval hit rate when temporal metadata was incorporated into the filtering process. That is not a marginal gain. It represents the difference between a system that reliably surfaces the correct document and one that gets there by chance roughly one in six fewer times.

This finding deserves attention because it reframes what "better retrieval" actually means. Much of the public conversation about improving RAG systems centers on embedding quality or chunking strategy. Those matter, but the temporal robustness research suggests that metadata-aware filtering can produce gains of a similar order of magnitude, often at a fraction of the computational cost of retraining or fine-tuning an embedding model.

Metadata Improves AI Accuracy and Reduces Hallucination

Hallucination remains the most persistent trust problem in generative AI, and much of the discourse around it treats the phenomenon as a model-level flaw to be solved through better training. But a meaningful portion of hallucination in RAG systems is a retrieval problem wearing a generation costume. If a system retrieves irrelevant, outdated, or low-authority material and passes it to the language model as context, the model will do exactly what it is designed to do: synthesize a fluent answer from what it was given. The result looks like a hallucination, but its root cause is upstream.

Work on reliable retrieval-augmented feature generation has examined precisely this relationship, showing that the reliability of downstream reasoning is closely tied to the quality of what gets retrieved in the first place. This is a useful corrective to any evaluation strategy that only scores final outputs. If a benchmark measures answer accuracy without also scoring retrieval precision and metadata fidelity, it risks rewarding systems that are lucky rather than sound, and it makes diagnosing failure far harder than it needs to be.

Provenance and Source Trustworthiness

Few domains illustrate the stakes of provenance more clearly than clinical decision support. A framework for auditable, source-verified clinical AI, integrating retrieval-augmented generation with explicit data provenance, was designed around a simple but demanding principle: every recommendation the system makes should be traceable back to a verifiable source, with that source's reliability made visible rather than assumed. This is provenance metadata doing real work, not as a compliance checkbox but as a functional requirement for a system operating in a high-stakes domain.

The broader lesson generalizes well beyond medicine. Any AI system that produces recommendations, summaries, or answers benefits from being able to answer the question "where did this come from" with something more specific than "the training data" or "a document somewhere in the index." Provenance metadata is what makes that specificity possible.

Temporal Reasoning: Knowing When Something Was True

Facts have expiration dates, even when the sentences describing them do not visibly change. A statement that was accurate in 2022 may be false today, and a document that has been formally superseded may still sit in a retrieval index looking exactly as authoritative as its replacement. The fuzzy time expression research referenced earlier addresses this directly, and it is worth returning to because temporal reasoning touches nearly every other capability discussed here. Explainability depends on knowing when a claim was made. Trust depends on knowing whether a claim still holds. Citation depends on being able to tell a user not just what a source says but whether it remains current. Metadata that encodes validity windows, supersession relationships, and publication dates gives a system the raw material to reason about all of this rather than presenting every retrieved passage as equally timeless.

Metadata Enables Explainable AI

Explainability is frequently discussed as though it were purely a matter of model interpretability: attention visualizations, feature attributions, and the like. But for retrieval-augmented systems, a more practical and arguably more useful form of explainability comes from metadata itself. If a system can show which document it drew from, who authored it, when it was published, and how authoritative that source is judged to be, it has already given a user most of what they need to evaluate the answer's credibility. Research introducing governance-driven agentic retrieval chains has pushed this idea further, proposing explicit evidence ledgers that log authority scores, temporal validity, and conflict resolution decisions across a multi-step retrieval process. This kind of ledger transforms explainability from a static disclosure into an auditable trail, which is a meaningfully different and more rigorous standard.

Metadata Supports Citation Generation

Citation is where metadata's value becomes most visible to an end user. A system that can point to the exact source of a claim, along with its date, authorship, and context, is offering something categorically different from a system that simply asserts a fact. Work applying retrieval-augmented generation to art provenance research within a major cultural heritage index demonstrates this well. In a domain like art history, where the chain of ownership and attribution is often the entire substance of the inquiry, metadata is not a supporting feature. It is the subject matter itself. The success of that research in surfacing accurate, explainable provenance chains for historical artworks offers a useful proof of concept for citation generation in far less specialized domains.

Metadata Helps AI Understand Document Relationships

Individual documents rarely stand alone. They respond to one another, build on one another, and sometimes contradict one another. Research on adaptive information management for retrieval-augmented generation has explored how systems can maintain coherent reasoning across multiple retrieval steps by tracking these relationships in working memory rather than treating each retrieval as an isolated event. This has particular relevance for historical research, where understanding how one document relates to, revises, or is revised by another is often the actual analytical task. Metadata that captures these relational links allows an AI system to support that kind of layered inquiry rather than flattening it into a series of disconnected facts.

Metadata Enables Better AI Evaluation

Evaluation of AI systems has tended to focus heavily on final-answer accuracy, but a growing body of work argues for scoring the retrieval and reasoning process itself. Research on self-evaluation driven strategy optimization in agentic retrieval introduced a self-assessment step in which the system evaluates the quality of retrieved evidence before generating a final answer. This kind of built-in scoring rubric, applied to metadata quality and source reliability rather than only to output fluency, represents a more rigorous approach to benchmarking. It suggests that future evaluation standards for RAG systems should include explicit metrics for provenance completeness, temporal accuracy, and citation traceability, not only for the correctness of the final generated text.

Metadata Improves Security and Governance

Metadata also does quiet but essential work in access control. Permission-aware retrieval, in which a system respects who is allowed to see which documents based on classification, ownership, or sensitivity metadata, is what makes AI systems viable in enterprise and regulated environments. The governance-driven framework mentioned earlier extends this idea by building authority scoring and audit trails directly into the retrieval chain, treating governance not as a separate layer bolted on top of the system but as something woven through its metadata architecture from the start.

The Future: Metadata-Aware AI Agents

Looking ahead, the most interesting frontier is not a smarter language model but a more metadata-literate one. Future agentic systems will likely reason over metadata with the same deliberateness they currently apply to text: weighing provenance, tracking temporal validity, assigning confidence scores, generating citations automatically, integrating with knowledge graphs, handling multimodal metadata across images, audio, and video, and preserving chain of custody across long, multi-step workflows. This shift would move AI systems away from merely producing plausible-sounding text and toward producing answers that are transparent, traceable, and genuinely defensible under scrutiny.

That is a meaningful distinction, and it is worth sitting with. Plausibility and defensibility are not the same standard. A fluent answer can be plausible and still wrong. A defensible answer, grounded in well-structured metadata, gives a user the means to check it. As AI systems take on more consequential tasks, from clinical support to historical research to enterprise decision-making, that difference is likely to matter more, not less. The path toward more trustworthy AI runs directly through the unglamorous, essential work of getting metadata right.

---

Sources

Chen, L.-C., Chen, H.-W., & Chen, M.-S. (2026). DeFuzzRAG: Handling Fuzzy Time Expressions for Temporal Robustness in Retrieval-Augmented Generation. Proceedings of AAAI-26. https://ojs.aaai.org/index.php/AAAI/article/view/40276

An Auditable and Source-Verified Framework for Clinical AI Decision Support: Integrating Retrieval-Augmented Generation with Data Provenance. Frontiers in Artificial Intelligence (2026). https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1737532/full

Reliable Retrieval-Augmented Feature Generation with Large Language Model Reasoning. Knowledge and Information Systems (Springer), 2026. https://link.springer.com/article/10.1007/s10115-026-02792-4

A ReAct- and RAG-Based Framework for Metadata Generation and Access in Relational Data Warehouse Processes. Big Data and Cognitive Computing (2026). https://doi.org/10.3390/bdcc10060172

LedgerRAG: Governance-Driven Agentic Chain of Retrieval for Dynamic Knowledge Scenarios. Electronics (MDPI), 2026. https://www.mdpi.com/2079-9292/15/7/1376

Reasoning with Memory: Adaptive Information Management for Retrieval-Augmented Generation. Findings of ACL 2026. https://aclanthology.org/2026.findings-acl.1834

Reflective RAG: Self-Evaluation Driven Strategy Optimization in Agentic Retrieval-Augmented Generation. Findings of ACL 2026. https://aclanthology.org/2026.findings-acl.648

Retrieval-Augmented Generation for Natural Language Art Provenance Searches in the Getty Provenance Index (2026). https://eprints.whiterose.ac.uk/id/eprint/238115

Ten Common Hallucinations in Genealogical AI

Ten Common Hallucinations in Genealogical AI

by Alvin Blackshear | Historian & Researcher <ablackshear@gmail.com>

If you've started using AI tools to help build out your family tree, you've probably noticed something exciting: AI can search patterns, suggest connections, and summarize records faster than any human researcher. But you may have also noticed something unsettling. Sometimes the AI is simply wrong, and it's wrong with total confidence. It doesn't hedge. It doesn't say "I'm not sure." It states fabricated facts as if they were pulled straight from a courthouse archive.

This isn't a quirk specific to genealogy tools. It's a well-documented feature of how large language models work. Researchers have found that AI systems are, in a sense, rewarded for guessing rather than admitting uncertainty, much like a student on a hard exam who writes down a plausible-sounding answer rather than leaving the question blank. The same dynamic that produces fabricated legal citations in court filings and nonexistent references in academic papers also produces invented ancestors, false marriages, and imaginary historical explanations in your family tree.

For beginning genealogists who are comfortable with AI tools but new to the discipline of genealogical research, this creates a real risk. Family history has its own standards of evidence (rules developed over generations of professional practice), and AI, for all its fluency, doesn't inherently know or follow them. Below are ten (plus a bonus eleventh) hallucinations you're likely to encounter, along with how to spot each one.

#1. Invented Ancestors

The most obvious and probably most common hallucination is the flat-out invented person. If your records show John Smith born in 1846, married to Sarah Jones, and died in 1912, an AI tool may cheerfully add "William Smith, father of John" to your tree, simply because a father-shaped gap exists and the model has learned that fathers usually appear in these slots.

No birth record names him. No probate record mentions him. No document anywhere establishes his existence. He is, in the most literal sense, a statistical guess dressed up as a fact.

How to catch it: Ask the AI directly, "What specific document supports this person's existence?" If it can't name one, or names one that turns out not to exist (see #5), strike the individual from your tree.

#2. False Parentage

This is a subtler and more dangerous error because it doesn't invent anyone; it just connects the right person to the wrong family. Imagine two men named James Brown, both born in Virginia in 1818, but belonging to two entirely unrelated families. An AI tool, pattern-matching on name, place, and approximate date, may merge them, attaching one man's documented parents to the other man's documented children.

Because online family trees copy from each other constantly, an error like this doesn't stay contained. It can spread across dozens or hundreds of public trees within months, becoming "true" simply through repetition.

How to catch it: Whenever two individuals share a name, county, and rough birth year, ask the AI to list the specific evidence distinguishing them: land records, church parishes, associates named as witnesses, distinguishing middle names. If it can't distinguish them, you probably have two people, not one.

#3. Collapsing Two People into One

Related to false parentage, this hallucination happens when the AI treats "same name, similar era, same county" as sufficient proof of "same person." Common names are especially vulnerable: John Williams, William Johnson, Mary Jones, Elizabeth Brown. In many rural communities of the eighteenth and nineteenth centuries, several unrelated people carried the same name at the same time.

How to catch it: Before accepting that two records refer to one individual, ask what independent evidence (land transactions, church membership, military service, or family testimony) ties them together, not just shared name and geography.

#4. Splitting One Person into Several

The mirror-image error: one real person recorded inconsistently, as "Robert," "Robt.," "Bob," and "R.J.," gets treated by the AI as four different people. Suddenly your family tree has artificially multiplied, with duplicate siblings or duplicate spouses cluttering the record.

Nineteenth-century clerks were not consistent spellers, and names were frequently abbreviated, translated, or anglicized. AI models, trained to treat text strings literally, can miss the obvious inference that a human researcher would make instantly.

How to catch it: When a "new" individual appears with a name variant of someone already in your tree, in the same location and timeframe, check whether the records could plausibly describe one person before treating them as separate.

#5. Invented Sources

Perhaps the easiest hallucination to catch once you know to look for it, and one of the most dangerous when you don't. AI tools will sometimes cite a source that sounds entirely plausible: "County Probate Book 14, page 327" or "Virginia Marriage Register, Volume 8." The citation has the right shape, the right format, the right level of specificity. It just doesn't exist.

This mirrors a broader, well-documented problem. A 2026 analysis of AI-generated academic writing found a sharp rise in nonexistent references following widespread adoption of large language models, with a conservative estimate of nearly 147,000 hallucinated citations appearing in 2025 alone. Similar patterns have shown up in legal filings, where courts have caught lawyers submitting briefs built on cases that were never decided because an AI tool invented them. Genealogy is not immune to the same failure mode; if anything, it may be more vulnerable, since archival citation formats are formulaic and easy for a model to imitate convincingly.

How to catch it: Never accept a citation at face value. Look the source up yourself, in the actual archive, library catalog, or online record collection, before recording it as fact. If you can't locate it, assume it doesn't exist until proven otherwise.

#6. Imaginary Historical Context

AI loves to explain. "The family likely moved west because of the Panic of 1837." It's a tidy, plausible sentence. It may even be historically reasonable. But plausible is not the same as documented, and an AI tool rarely marks the difference clearly. Historical interpretation, offered without evidence specific to your family, quietly becomes historical fiction wearing the authority of fact.

How to catch it: Treat any sentence containing words like "likely," "probably," or "given the context of the time" as a hypothesis to test, not a fact to record. Ask what document (a land sale, a migration record, a letter) actually supports the claimed motive.

#7. Overconfident Translation

If your ancestors left records in German, Polish, Latin, French, or Dutch, AI translation can feel like a superpower, until it isn't. AI models translate old church and civil records with impressive fluency, but that fluency can mask serious errors: invented words to fill illegible gaps, misread names, or entire phrases reconstructed from context rather than actually read. Genealogists working with foreign-language parish records have reported AI tools fabricating names, places, and even whole narrative details that don't exist in the original document.

How to catch it: For any translation affecting a name, date, or relationship, ask for a word-by-word rendering alongside the smooth translation, and compare it against the original image yourself, or have a human speaker of the language check ambiguous passages.

#8. Fabricated Relationships

A record lists a witness, Samuel Jones, at a wedding or a will signing. The AI concludes he must be an uncle, brother, father, or cousin. But witnesses in historical documents were frequently neighbors, business associates, fellow church members, or simply friends, not relatives at all. The AI's eagerness to complete the family picture leads it to assign a relationship where the record specifies none.

How to catch it: Treat every unlabeled name in a record (witness, executor, sponsor, bondsman) as a relationship of unknown type until independent evidence establishes otherwise.

#9. Timeline Repair

Large language models dislike contradictions, and historical records are full of them. If a record shows a birth in 1820 and a marriage in 1832, and something else nearby seems inconsistent, an AI tool may quietly adjust one of the dates so everything appears internally consistent. The apparent contradiction vanishes, but so does the actual historical evidence, replaced by a smoothed-over fiction that fits the model's sense of what "should" be true.

How to catch it: When a date changes between one AI response and another, ask why, and check both figures against the original record. Real historical records are often genuinely inconsistent, and that inconsistency is itself information, not a bug to be silently fixed.

#10. Source Laundering

This is probably the least recognized hallucination, and arguably the most insidious. AI tools trained on the vast ocean of user-submitted online family trees absorb whatever errors already exist there, including all the mistakes described above, made by human researchers over the past two decades. The AI then presents those errors back to you as though they were settled historical fact.

Nothing here is invented from scratch. That's what makes it dangerous. The information already circulates online, attached to seemingly credible trees and family history websites, so it carries an unearned appearance of authority. The AI hasn't lied exactly; it has simply repeated community folklore as though it were consensus.

How to catch it: Be especially skeptical of any claim the AI attributes to "commonly accepted" genealogy or "widely documented" family history. Ask what primary source underlies the claim, not what other online trees say about it.

An Eleventh Hallucination: Certainty Inflation

There's one more pattern worth naming, even though it's rarely discussed: certainty inflation. Professional genealogists are trained to say, when the evidence doesn't fully support a conclusion, "the evidence is insufficient." AI tools are much less comfortable with that kind of humility. Instead, they tend to write things like "based on available evidence, John was almost certainly the son of..."

That phrase, "almost certainly," can represent an enormous leap well beyond what the underlying evidence supports. This isn't accidental. Research into why language models hallucinate has found that the training and evaluation processes used to build these systems reward confident, plausible-sounding answers over honest expressions of uncertainty, the AI equivalent of a student guessing on a test rather than leaving a blank. The result is a model that is statistically inclined to sound more certain than the facts warrant.

Toward a Genealogical AI Evaluation Framework

None of this means AI is useless for family history research; far from it. It means AI output needs to be treated the way a careful historian treats any single source: as a claim to be verified, not a fact to be recorded.

One useful discipline is to treat every AI-generated genealogical statement as a research hypothesis, never as evidence in itself. Accept only what can be traced back to an identifiable primary or reliable secondary source, and preserve uncertainty wherever the historical record itself remains uncertain. In practice, that means running each AI claim through a short mental checklist:

Criterion	Key Question
Source authenticity	Is every factual claim traceable to a real, locatable source?
Evidence chain	Can the conclusion be reconstructed from the cited evidence?
Identity confidence	Has the AI distinguished between similarly named individuals?
Chronological consistency	Are dates plausible without silent correction?
Geographic plausibility	Do locations fit historical boundaries and migration patterns?
Uncertainty disclosure	Does the AI distinguish evidence from inference?
Genealogical Proof Standard alignment	Would the conclusion satisfy professional genealogical standards?

Applying a framework like this won't eliminate hallucinations (no current AI system is free of them), but it will keep you from mistaking a fluent, confident answer for a well-documented one. AI can be a genuinely powerful research assistant for genealogy: it can suggest search strategies, summarize long documents, translate unfamiliar scripts, and spot patterns across large record sets faster than any human could alone. Just remember that its job is to help you find the evidence, not to become the evidence itself.

________________________________

Sources:

Kalai, Adam, et al. "Why Language Models Hallucinate." arXiv, 4 Sept. 2025, https://arxiv.org/abs/2509.04664.

Newsham, Jack. "AI Hallucinations in Court Documents Are a Growing Problem, and Data Shows Lawyers Are Responsible for Many of the Errors." Business Insider, 27 May 2026, https://www.businessinsider.com/increasing-ai-hallucinations-fake-citations-court-records-data-2025-5.

Yin, Yian, et al. "LLM Hallucinations in the Wild: Large-Scale Evidence from Non-Existent Citations." arXiv, 8 May 2026, https://arxiv.org/abs/2605.07723.