AI for Genealogy: The Complete Guide to Using AI for Family History Research
AI for genealogy: which tools work, where they fail, and how to build a workflow that finishes your family's story. Tested on real records.
AI for genealogy is the use of artificial intelligence tools — Claude, ChatGPT, Gemini, Perplexity, and others — to research ancestors, read historical records, organize evidence, and write family history stories that people actually want to read.
This guide covers what AI can do for family historians right now, which tools work best for which tasks, where AI fails, and how to build a workflow that moves your research toward a finished story instead of an endless collection of records.
Everything here is tested on real genealogy work. Real Pennsylvania records, real ancestors, real writing projects. No hypotheticals.
What AI actually does for family history research
AI doesn't search databases for you. Ancestry, FamilySearch, and MyHeritage already do that. What AI does is work with the records you've already found.
The distinction matters. A search engine retrieves what exists. An AI generates what fits — summaries, timelines, context, draft prose, analysis of what a record means in the context of your ancestor's life.
Here is what that looks like in practice:
Reading old handwriting. You photograph a deed from 1782 and upload it. The AI reads the handwriting and gives you a transcription. Not a perfect one — you verify it against the original — but a starting point that saves hours of squinting at faded ink.
Understanding what a record means. Census records, probate inventories, tax lists, and land deeds all have specific historical context that changes how you interpret them. AI tools can explain what a census enumerator was told to write in 1870 versus 1880, why the ages don't match, and what that tells you about your ancestor.
Organizing scattered research. You have twenty years of documents across binders, hard drives, and browser tabs. AI can process a folder of research notes and produce a chronological timeline, a gap analysis showing what you know and what you don't, and a structured summary with citations.
Writing family stories. This is where AI changed everything for family historians. Not writing the story for you — that produces generic prose nobody wants to read. But drafting timelines, building outlines, suggesting narrative structures, and producing scaffolding you rewrite in your own voice. The difference between staring at a blank page and staring at a rough draft you can fix is the difference between finishing and not finishing.
Building research plans. You tell the AI what you know about an ancestor — name, dates, locations, records already found. It identifies record types you haven't checked, repositories you haven't visited, and questions your existing evidence doesn't answer.
Which AI tools work best for genealogy
Not every AI tool does the same thing well. After testing Claude, ChatGPT, Gemini, and Perplexity on real genealogy research across hundreds of sessions, here is what each one does best.
Claude (by Anthropic)
Claude is the AI most family historians end up staying with for research and writing once they try it. Its strengths are long-document reading, careful writing, and the ability to hold complex genealogy projects in memory across sessions.
Claude's Projects feature lets you upload your research files and set custom instructions so every conversation starts with context. No more re-explaining who your ancestor is every time you open a new chat. Claude's Custom Skills let you save reusable instructions for specific tasks — transcription, blog post drafting, citation formatting — so the AI works the way you work.
Best for: reading handwriting, writing family narratives, building research plans, long document analysis, project-based research.
Read the full Claude for Genealogy guide →
ChatGPT (by OpenAI)
ChatGPT is the tool most people try first. It has the largest user base, the most name recognition, and a generous free tier. For genealogy, ChatGPT is strong at quick factual questions, historical context, and brainstorming. Its Custom GPTs let you build specialized assistants, though the setup is less flexible than Claude's approach.
Best for: quick research questions, historical context lookups, brainstorming narrative angles, creating Custom GPTs for specific tasks.
Read the full ChatGPT for Genealogy guide →
Gemini (by Google)
Gemini's standout feature for genealogists is its spatial grounding capability — the ability to understand where things are on a page. For handwritten historical records, this means Gemini can not only read the text but understand the layout: which column a number belongs to, where a margin note sits, how a table is structured. This is particularly valuable for census records, court dockets, and tax lists where position on the page carries meaning.
Best for: reading structured historical documents (census, tax lists, court records), image analysis where layout matters, integration with Google Workspace.
Read the full Gemini for Genealogy guide →
Perplexity
Perplexity is different from the others. It combines AI with live web search, citing its sources with links you can verify. For genealogy, this makes it useful for finding contextual information — what was happening in a specific county in a specific year, which archives hold records for a specific church, what a historical term means in context.
It is not the right tool for reading your documents or writing your stories. It's the right tool for answering factual questions about the historical world your ancestor lived in.
Best for: historical context research, finding repositories and archives, verifying facts, answering "what was happening in [place] in [year]" questions.
Read the full Perplexity for Genealogy guide
New to AI for family history? Take the AI Fundamentals 5-part course and learn the basics of how it works for genealogy research and writing
Where AI fails in genealogy (and how to catch it)
AI makes mistakes. Knowing how it fails is as important as knowing what it does well.
Confabulation
AI does not search databases. It predicts what text should come next based on patterns. When it doesn't have enough information, it fills the gap with something that sounds right but isn't. In genealogy, this means AI will sometimes invent dates, create nonexistent records, or confidently state facts that are wrong.
This is the most dangerous failure mode for genealogists because the output looks authoritative. The AI doesn't say "I'm guessing." It presents invented information with the same confidence as verified facts.
How to catch it: Never trust a date, record citation, or factual claim from AI without verifying it against an original source. If the AI says a record exists, find that record yourself. If it gives you a date, check it against your documents.
Mixing up people with similar names
AI struggles with same-name individuals in the same location and time period. If your ancestor John Smith lived in Lancaster County in 1830, and there were three John Smiths in the county, AI will sometimes merge information from all three into one person without telling you.
How to catch it: Be explicit about distinguishing details. Provide birth years, spouse names, and township-level locations. When the AI gives you information, ask it to identify which specific person the information applies to.
Over-interpreting records
AI can read what's on the page. It cannot know what's not on the page. A census record showing a family of four doesn't mean there were only four children — it means four were home on census day. AI sometimes draws conclusions from records that the records don't actually support.
How to catch it: Ask the AI to distinguish between what the record says and what it might mean. Treat AI interpretations as hypotheses to test, not conclusions to accept.
Read more: Should I Trust AI for Genealogy Research →
How to build an AI genealogy workflow that finishes stories
The gap most genealogists experience is not a technology gap. It's a workflow gap. You have records. You have AI tools. You don't have a process that moves from one to the other and produces a finished story.
Here is the workflow that works. This is the same process used inside Chronicle Makers, where 95% of Sprint participants finish a complete family chronicle in ten days.
Step 1 — Organize what you have
Before you open any AI tool, know what you're working with. Build a research summary for one ancestor: what records you have, what dates are confirmed, what gaps exist. AI can help with this step — upload a folder of documents and ask for a chronological inventory.
Step 2 — Scope your story
Choose one ancestor. Define the time period, the central question, and the angle. "The life of Great-Grandma Miller" is too broad. "How Anna Miller kept the farm running after her husband's death in 1918" is a story you can actually write.
This is what the STORI Method calls the Scope step. It's the most important decision in the entire process because it determines everything that follows.
Step 3 — Build context with AI
Use Perplexity or ChatGPT to research the historical world your ancestor lived in. What was happening in their county, their industry, their community during the years that matter? Feed this context into Claude alongside your research documents. Now the AI understands both your ancestor and their world.
Step 4 — Draft with AI, rewrite as yourself
Ask Claude to produce scaffolding — a rough timeline, section outlines, opening paragraph options. Not a finished story. Scaffolding is a starting point you tear apart and rebuild in your own voice. The AI removes the blank-page paralysis. You supply the voice, your perspective, and the family knowledge that makes it real.
Step 5 — Source everything
Every fact, every date, every record reference in your finished story gets checked against original sources. AI helped you write faster. Your job now is to make sure what you wrote is true.
The named methods behind this work
Chronicle Makers uses named frameworks that travel with you once you learn them.
The STORI Method — Scope, Thread, Originate, Reflect, Iterate. The five-step path from scattered research to a finished family chronicle. Learn the STORI Method →
Chronicle Readiness — How to know whether you have enough research to write about a specific ancestor, or whether you need one more thing first. Learn more →
The Full Pantry Problem — What happens when you have everything you need to write and still can't start. The most common stuck point in family history. Learn more →
The Finisher Framework — The identity shift from "researcher who hopes to one day write" to someone who finishes chronicles. Learn the Finisher Framework →
Getting started
If you're new to AI for family history, start with the free 5-Day AI Beginners Course. No tech background needed. By day five, you'll have used AI on your own records.
Start the Free 5-Day AI Course →
If you have research and you're ready to write, the Chronicle Makers community is where the work happens. Live sessions every week, the 10-Day Writing Sprint, and a room full of people who are actually finishing their writing.
Frequently asked questions
Is AI safe to use for genealogy? AI is a tool. Like any tool, it depends on how you use it. The main risk is trusting AI output without checking it against original sources. If you treat AI suggestions as hypotheses to test rather than facts to accept, it's safe. If you copy AI output into your family tree without checking it, you'll introduce errors.
Which AI tool should I start with? If you've never used AI, start with ChatGPT — it has a free tier and the lowest barrier to entry. If you're ready to do serious genealogy work, try Claude Pro. The Projects feature and Custom Skills make it the strongest tool for ongoing research and writing.
Can AI read old handwriting? Yes, with caveats. Claude and Gemini can both read historical handwriting with varying accuracy depending on the quality of the document. Results are good enough to save significant time, but you must check every transcription against the original.
Will AI write my family story for me? It can generate text, but the result reads like AI wrote it — generic, flat, and missing the family knowledge only you have. The effective approach is using AI for scaffolding (outlines, timelines, rough drafts) that you rewrite in your own voice. That's the difference between a story your family skims and one they read aloud at Thanksgiving.
How much does AI for genealogy cost? Free tiers exist for ChatGPT, Claude, and Gemini. Serious genealogy work benefits from a paid plan: Claude Pro, ChatGPT Plus, or Gemini Pro. One paid subscription to one tool is enough to start. Most genealogists don't need (or use) all of them.
What if I'm not a tech person? If you can type a question into Google, you can use AI for genealogy. The tools are designed for conversation — you type in plain English and get a response. No coding, no technical setup. Lab 1 inside Chronicle Makers exists specifically for people who don't think of themselves as tech people.
This guide is maintained by Denyse Allen, author of four books on Pennsylvania genealogy (two in the DAR Library permanent collection) and founder of Chronicle Makers. Last updated May 2026.
Related guides: Claude for Genealogy · ChatGPT for Genealogy · Gemini for Genealogy · Perplexity for Genealogy · Is AI Safe for Genealogy? · The STORI Method · The Finisher Framework