US Government Uses AI to Create Public Content

When your government posts a video online, how do you know if a human or AI created it?

The US Department of Homeland Security is using AI video generators from Google and Adobe to create and edit content shared with the public, according to a newly released document. This revelation comes as immigration agencies have flooded social media with content supporting President Trump's mass deportation agenda—some of which appears to be AI-generated.

The Government's AI Arsenal

Wednesday's document provides the first concrete inventory of commercial AI tools used by DHS. In a section about "editing images, videos or other public affairs materials using AI," it reveals that DHS uses Google'sVeo 3 video generator and Adobe Firefly, estimating between 100 and 1,000 licenses for these tools.

The agency also employs Microsoft Copilot Chat for generating document drafts and summarizing reports, plus Poolside software for coding tasks. Google'sFlow tool combines the Veo 3 video generator with filmmaking capabilities, allowing users to create hyperrealistic videos complete with sound, dialogue, and background noise.

Adobe launched its Firefly generator in 2023, promising it doesn't use copyrighted content in training or output. Like Google's tools, it can generate videos, images, soundtracks, and speech.

Immigration Content Under Scrutiny

Immigration and Customs Enforcement (ICE), part of DHS, has dramatically increased its social media presence as deportation operations expand across US cities. The agency has posted content celebrating "Christmas after mass deportations," referenced Bible verses, showed faces of arrested individuals, and shared recruitment ads.

Many of these videos appeared AI-generated, but until now, there was no concrete evidence about which AI models might be creating this content. The agencies have also repeatedly used music without artist permissions in their videos.

Verifying which company helped create specific content remains impossible. While Adobe offers "watermark" options to disclose AI generation, these markers don't always survive when content is uploaded and shared across different platforms.

Tech Worker Resistance

More than 140 current and former Google employees and over 30 from Adobe have pressured their employers to take a stance against ICE activities in recent weeks. Google's leadership has remained silent on the issue.

This isn't the first time tech companies have faced scrutiny over government partnerships. In October, Google and Apple removed apps designed to track ICE sightings from their app stores, citing safety risks.

Broader AI Government Use

The document also revealed DHS uses facial recognition apps and other niche AI products, indicating the agency's AI adoption extends far beyond content creation. This comprehensive AI integration raises questions about transparency, accountability, and the public's right to know when they're consuming AI-generated government communications.