This isn’t ‘unedited raw security footage’ of the White House correspondents’ dinner shooter
PolitiFact – Rulings and Stories — 2026-04-29 15:59:00 — www.politifact.com
Confusion and Misinformation Following White House Correspondents’ Dinner Shooting
In the aftermath of a shooting at the April 25 White House Correspondents’ Association dinner, social media became a hotbed of confusion, exacerbated by the use of artificial intelligence in video analysis. President Donald Trump posted a low-quality security video on Truth Social, showing a suspect, later identified as Cole Tomas Allen, rushing through a security checkpoint at the Washington Hilton hotel. This video, captured as Secret Service agents pursued the suspect, was the original unedited footage related to the incident.
However, the situation became muddled when some users employed AI to enhance the clarity of the video. This edited version was shared widely, including by conservative commentator Benny Johnson on X (formerly Twitter), who initially did not disclose the AI enhancements. The AI alterations introduced several inaccuracies, such as morphing agents’ hats and adding indistinct shapes, leading to further confusion among viewers.
Irregularities in AI-Enhanced Footage
The AI-enhanced video, first modified by X user "Seth Weathers," contained noticeable discrepancies not present in the original footage:
- Two agents appeared to be kneeling down in an unrelated action to the suspect’s movement.
- An agent’s cap transformed into a beanie in subsequent frames.
- A large white box overlaid on the suspect vanished post-security checkpoint.
- The lettering on the agents’ uniforms did not correspond with any known Secret Service division.
- A blurry shape intermittently resembled furniture or a kneeling agent in formal attire.
These irregularities led to the video being mistakenly represented as "unedited raw security footage," a claim that has been debunked.
What’s Unclear / What to Watch
- The full extent of the AI modifications and their source remains unclear.
- Further clarifications from users who shared the edited footage might emerge.
How This Sits Against Accuracy, Norms, and Governing Rules
Truth and Evidence
The incident underscores the complexities and potential pitfalls of using AI in critical contexts like security footage analysis. Normally, credible claims about such footage would require verification through independent sources or confirmation from involved security agencies. In this case, the alterations made by AI were acknowledged by the user who edited the video, providing some level of transparency about the source of confusion.What the Excerpt Shows About Verifiable Lies
The claim that the video was "unedited raw security footage" is directly contradicted by the evidence within the video itself, as noted by the discrepancies listed. This claim, therefore, can be assessed as false based on the content of the excerpt.This episode serves as a cautionary tale about the reliability of digital content and the critical need for clear communication regarding any modifications made to potentially sensitive materials.