ICE Agent Self-Pepper-Spray Video Revealed as Fake

In early October 2025, a video depicting an Immigration and Customs Enforcement (ICE) agent accidentally pepper-spraying himself circulated widely across social media platforms, generating thousands of shares and comments. The footage, however, has been confirmed as an artificial intelligence fabrication.

The 10-second clip showed what appeared to be an ICE agent discharging pepper spray near protesters, only to have the aerosol blow back into his own face. A voice off-camera remarked, “Watch the wind, man,” followed by “He got himself,” accompanied by laughter. The unseen commentator then asked, “You all right, officer?” to which the agent replied, “I’m good. I’m good.” The voice concluded with, “Karma’s quick, dude.”

The video gained particular traction amid heightened tensions surrounding immigration enforcement. The Trump administration had recently intensified deportation operations and attempted to deploy National Guard troops to several major cities including Portland and Chicago—a move that was temporarily blocked by a federal judge in Illinois.

One user who shared the clip on X (formerly Twitter) captioned it: “This ICE agent ended pepper spraying him and it serves him right! It’s obvious they have no clue what the hell they’re doing!” This post alone accumulated substantial engagement before fact-checkers identified the content as fake.

Several telltale signs revealed the video’s artificial origins. Most prominently, the footage contained a watermark for a TikTok channel with “AI” in its handle. The original TikTok post was captioned “Portland ice protest backfired,” but close inspection revealed multiple inconsistencies typical of AI-generated media.

The timing of the video’s appearance coincided with OpenAI’s release of Sora 2, an advanced text-to-video generation tool that, for the first time, included synchronized audio capabilities. This release triggered a surge of deceptively realistic AI videos across social platforms, with this ICE agent footage representing one example of the technology’s increasingly convincing output.

Technical analysis revealed numerous flaws that confirmed the video’s artificial nature. The pepper spray appeared to emanate from both of the agent’s hands, despite him only holding a canister in his right hand. When he raised his left hand, it was clearly empty. Additionally, the background featured nonsensical text reading “OF DEI LBNM EPARTMENT”—a common issue with AI generators that struggle to render coherent text within images.

Further analysis showed the agent’s facial movements failed to synchronize properly with his dialogue, another limitation of current AI technology. OpenAI has stated that all content generated through their platform carries visible watermarks specifically to distinguish AI creations from authentic footage.

The circulation of this fake video occurs amid growing concerns about AI’s potential to spread misinformation, particularly during politically charged moments. Immigration enforcement activities have been a flashpoint in American politics, making fabricated content in this domain particularly potent for viral spread.

Media literacy experts note that as AI-generated content becomes increasingly sophisticated, viewers must approach visual media with heightened skepticism, particularly when content aligns too conveniently with political narratives or depicts emotionally charged scenarios.

This incident follows other documented cases of AI-generated media being mistaken for authentic footage. Another recent example involved a fake video purportedly showing a protester being pepper-sprayed by a soldier who referred to himself as “Sergeant Pepper,” which was similarly created using OpenAI’s Sora technology.

As AI tools become more accessible and their outputs more convincing, the challenge of distinguishing fact from fiction in visual media continues to grow more complex for social media users, journalists, and the general public alike.