Detect Deepfakesby Resemble AI
Deepfake case study · Image

MAGA 'Beautified' America's Founders With AI — And Historians Say The Images Reveal A Troubling…

Historians and critics raise concerns over AI-generated portraits on the Freedom 250 website that inaccurately depict American Revolutionary figures

Incident date
Jul 2024
Target
Abigail Adams
Updated Jul 3, 2026 · 2 min read

In July 2024, the MAGA-affiliated Freedom 250 website launched a series of AI-generated portraits depicting figures from the American Revolutionary War. The images sparked significant backlash from historians and online users who criticized the depictions for their lack of historical accuracy and the use of heavy beautification filters.

What happened

The portraits featured on the Freedom 250 website were generated using artificial intelligence, resulting in images that historians described as homogenous and inaccurate. A primary point of contention was the rendering of Abigail Adams, which social media users likened to a Disney princess, noting that it bore no resemblance to verified portraits painted from life. Critics labeled the process as yassification, a term referring to the application of excessive beauty filters.

Beyond the specific depiction of Abigail Adams, experts identified several systemic issues with the AI-generated collection. Historians pointed out that the figures appeared strikingly similar to one another, wearing near-identical blue clothing that lacked historical grounding. Kari Winter, a professor at the University of Buffalo, characterized the images as a form of AI-generated propaganda that prioritizes a modern, beautified aesthetic over historical reality. She argued that these portraits function to reinforce specific ideologies by suggesting that historically significant women must adhere to a cookie-cutter standard of beauty.

Similarly, male figures in the collection were noted for their interchangeable, cut-and-paste appearance, which experts argued erased the gritty specificity of real historical individuals. Furthermore, the AI struggled with period-appropriate details; for instance, the clothing in the portraits was compared to costumes from the Gilded Age rather than the 18th century. Experts emphasized that because AI models are often trained on modern datasets, they lack the nuance to distinguish between different historical eras without highly specific prompting. Historians expressed concern that presenting these AI-generated images as scholarly and accurate content risks spreading misinformation and sidelining the actual contributions of women and marginalized people in the founding of the United States.

Sources