Artificial Intelligence has been with us for a while now and it’s definitely here to stay. Actually, scientists have worked on the concept for decades, but it’s only been recently released to the masses for general use.

Genealogists have begun embracing various aspects of AI, although in the beginning, without even realizing it. AI has developed to the point today where it’s time to examine its pros – of which there are many – but also some of its cons.

Remember how excited we all were when OCR was used to index digitized historical newspapers?

More recently, MyHeritage used AI to enable us to add movement and computerized male and female voices to our family photos.

Now, AI can provide background information on subjects in photos, estimating when they were taken and, in some cases, where.

Computer assisted transcription and translation of documents was a huge boon for genealogists as it saved us all so much time and energy, especially when the records were in another language.

Long articles can be summarized in a few seconds.

Websites such as Perplexity, ChatGPT 4o and others can answer historical questions, locate resources and even write articles within parameters that we specify.

Let’s also not forget FamilySearch full-text search, launched via U.S. land and probate records and is ever-expanding.

AI can even, at times, guide researchers to new possible resources that might have been overlooked.

As long as the human researcher remembers that AI is very capable of responding with lies (which some call ‘hallucinating’) and he/she follows through with checking and verifying information, Artificial Intelligence is offering many pros in its uses.

However, AI definitely has some cons, too.

First, let’s look at AI-generated content, whether it be for genealogical purposes or not. Professional writers clearly state when part or all of an article or story has been computer-generated.

For the moment, let’s equate those writers with genealogists who do their own research, carefully following clues and proving points with the best evidence possible at the moment.

Now, for the con – think about all those online trees with massive errors in them which proliferate like rabbits because they are copied by name collectors who don’t verify anything.

If we’ve been surprised at how quickly genealogy mis-information multiplied when we made the leap from society newsletters and journals to online trees, just imagine how fast incorrect digital information will multiply by researchers who spread AI content (without stating it as such) AND they don’t bother to fact check.

Accessing poorly documented or undocumented family information will present the serious researcher with many more “facts” to prove, disprove and accurately cite.

Another con, and this is my opinion and has become a pet peeve, is the use of AI-generated images.

Those images aren’t a 100% negative in my eyes.

For example, using AI images with AI-created poems, songs and biographies is a fun way to look at and share family history, especially if no actual photos of the subject are available.

Another way to use AI images is to create a branding logo or icon to draw attention to a specific topic or business. Since, as of 2025, the government has announced that computer created images cannot be copyrighted, writers can easily add logos to their work at no cost.

Those are both very appropriate uses of AI-generated images.

Now for the con. Again, this is my own opinion here so readers are certainly free to disagree.

I don’t understand why serious researchers and writers are adding AI images to their hard work. To me, it’s like adding cartoons to scholarly works. Those images don’t add anything to the quality of the post, article or story.

In fact, I believe AI images DEVALUE the writer’s final product. It also tends to make me think that the writer was too lazy to find an actual photo that would complement the work.

Here are a few examples:

Why add an AI image of a family and cabin on homesteaded land in the 1800s when copyright free images of real cabins can be accessed on the Library of Congress website and many other libraries with online digital collections?

Why add an AI image of a person sitting at a table reading a piece of paper (e.g. a will) when real photographs of that particular county courthouse are online, again copyright-free, and would greatly enhance the story by giving the reader a visual image of the actual building?

Why use an AI image of a library reading room when, again, there are many photos to be found online that can depict the same scene?

It really pains me to see otherwise excellent article and stories that have been cluttered with AI images. Even one in a non-fictional work is one too many.

I also have to admit that I actually cringe and have started to pass over (as in delete from feedly) serious articles when I see an AI image in them. I would much rather read an article without any images than one with AI content added.

Think about it – when was the last time you saw an AI-generated image in The Register (published by the New England Historic Genealogical Society)? Or in a case study published in a society newsletter? Or an any product meant to convey scholarly work? Answer: NEVER.

I really hope that the fascination of adding computer-created images wears off soon and the its use is relegated to the fun, non-professional genealogically related activities where it belongs.

Rant done!