On Aug. 11, the Museum of Democracy opened “Pathway to the Presidency: Road to the White House” on the lower level of the University of Arizona Campus Store. The exhibit “explores the history of American democracy through the creative campaign tools that shaped it—flags, posters, parade lanterns, stickers, games, and more.” It is supported by a grant from the Arizona Board of Regents and pieces from the Wright Family Collection.
Alongside the heaps of political merchandise, one of the spotlight features of the exhibit is a kiosk through which you hear from “these multilingual, AI-powered agents [which] represent a significant leap forward in how educational institutions can bring history to life for their audiences.”
This is a flowery way of saying that the Museum of Democracy has installed a large tablet screen displaying an AI-generated moving portrait of George Washington. This robot responds with “historically grounded answers, creating a personalized and immersive educational experience based on what interests the visitor.”
When I first saw this via an Instagram post on the University of Arizona’s page, I was taken aback and also reminded of another case of AI mimicry.
In 2024, Switzerland’s oldest church, St. Peter’s Chapel, debuted “Deus in Machina,” developed by a team at the Immersive Realities Research Lab at the Lucerne University of Applied Sciences and Arts. Users entered a confessional booth (which should be noted was explicitly stated not to replace the radiation of confession) and spoke to an AI-generated version of Jesus Christ.
I believe the word for this is blasphemy, among other things.
While it was made clear that users weren’t actively talking to Jesus himself, we’ve all seen an older relative repost a dog doing BMX tricks on a three-wheeled motorbike enough times to know that people don’t often see what’s right in front of them.
But what’s the harm?
Obviously, in the case of “Pathway to the Presidency”, the AI is, hopefully, highly vetted to avoid spreading misinformation, but this isn’t always the case. Every day, thousands of artificially generated videos are uploaded to social media, all in varying quality. I’m sure many remember the videos of bunnies and bears jumping on trampolines that people insisted were real. In the same vein, there are many videos of historical figures that are very, very convincing. While some of the videos are silly, like George Bush singing “Beautiful Girls” by Sean Kingston, others are intentionally created to spread disinformation. For example, in August 2025, TV newscaster Chris Cuomo caught flack for falling for an obvious AI video of Congresswoman Alexandria Ocasio-Cortez talking about Sydney Sweeney’s ad campaign on the floor of the House of Representatives. This did not happen of course, it was clearly labeled AI, but he re-shared it with a snarky quip to his millions of followers all the same. Whether or not it’s just for clicks and views, some of the opinions espoused by these AI mockeries of high profile figures can create real confusion.
Any AI-generated depiction of a historical figure, no matter how well disclosed and vetted, helps desensitize us to AI-generated content in general. It’s teaching those who are coming to learn about American democracy that it’s ok and good to learn from AI-generated photos and videos. But not only is it harming our perception of the world, but it’s also affecting our view of the historical figures we’re imitating. Sure, we know Thomas Jefferson isn’t actually talking to us, but if he were to go rogue and start cursing out a third grader, that would be the first thing they would think about when they hear “Thomas Jefferson” for the rest of their lives. Furthermore, an AI Thomas Jefferson’s accuracy to the real historical figure wouldn’t be acceptable. Jefferson owned slaves, so if an Al Thomas Jefferson responded with his genuine views, it would be socially unacceptable, and if it responded with fake, flowery views, it is not a depiction of the man himself, flaws and all. When we posthumously put words in the mouths of real people, we dilute the power, for good or bad, of the words they said themselves.
The AI display in the “Pathway to the Presidency” exhibit isn’t a massive issue in the grand scheme of things; however, it is indicative of the broader normalization of generative AI in our lives. AI has no place in academia at any level, whether used by teachers or students. By spoon-feeding learners historical facts in this way, we undermine their ability to think critically about contemporary issues. Seeing and hearing is not always believing, and in the future, I hope the University of Arizona will avoid promoting this type of AI.
