DHS Agents Use Meta Smart Glasses & AI for Immigration Surveillance

The lines between consumer technology and government surveillance are increasingly blurring, raising profound questions about privacy, civil liberties, and the rapid integration of AI into daily life. Recent revelations from 404 Media highlight a disturbing trend, from federal agents leveraging off-the-shelf smart glasses to the shadowy world of AI-driven voice replication and the unauthorized use of ubiquitous license plate reader networks.

One of the most immediate concerns centers on the sighting of a Customs and Border Protection (CBP) officer wearing Meta Ray-Ban smart glasses during an immigration enforcement action in Los Angeles in June 2025. This incident immediately sparked debate over the creeping presence of consumer-grade AI eyewear in federal policing operations and its potential impact on Department of Homeland Security (DHS) biometric identification systems. While Meta’s Ray-Ban glasses are marketed with a camera, microphones, and live-streaming capabilities, the company maintains they do not come equipped with built-in facial recognition. However, experts note it is “trivial” to funnel the glasses’ video feed to a separate device and apply third-party face-matching software in near real-time, a capability not sanctioned by CBP. Internal DHS sources confirm there is no official policy specifically addressing Meta Ray-Ban smart glasses, and existing regulations explicitly forbid on-duty personnel from using personal recording devices for official enforcement actions, particularly those involving First Amendment-protected activities without reasonable suspicion. Privacy advocates, including the ACLU, view such unauthorized use of cameras as potentially intimidating, aligning with behaviors designed to “terrorize people,” and underscoring a growing disregard for the boundary between personal tech and state surveillance. This incident follows a chilling precedent earlier in the year when a terrorist in New Orleans reportedly used Meta Ray-Ban smart glasses for reconnaissance before a deadly attack, demonstrating their capacity for discreet recording.

Meanwhile, the entertainment industry grapples with its own technological reckoning as voice actors face a complex bargain with artificial intelligence. AI has revolutionized voice production, offering rapid, cost-effective, and versatile solutions through sophisticated text-to-speech technologies that can mimic human nuances. This shift has led to over half of content creators incorporating AI voices into various media, driven by efficiency and cost savings, even though a significant majority (73%) still express a preference for human narration due to its irreplaceable emotional depth and spontaneity. In response to this evolving landscape, the Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) has entered into landmark agreements with AI startups like Replica Studios and Narrativ, allowing union members to license their digital voice replicas for use in advertisements and video games. These deals aim to provide actors with control over their digital likenesses and ensure fair compensation for their AI-generated voices. Yet, a darker side persists, with some voice actors suing AI companies for alleged unauthorized voice cloning. Industry projections suggest a potential 30-50% reduction in traditional voice acting jobs within the next decade, transforming a multi-billion-dollar market as AI models, trained on licensed human voices, become increasingly sophisticated and reduce the need for human talent.

Concurrently, a troubling pattern of unauthorized access and data misuse has emerged within the realm of public safety technology, specifically involving Flock Safety’s ubiquitous Automatic License Plate Reader (ALPR) systems. These cameras, installed across thousands of communities, continuously scan and record vehicle movements, creating vast databases searchable by law enforcement, often without a warrant. The “hacking” highlighted by 404 Media refers not to a malicious breach, but to federal agencies like the Drug Enforcement Administration (DEA) and Immigration and Customs Enforcement (ICE) gaining de facto access to Flock’s network by leveraging credentials of local police departments, effectively circumventing direct federal oversight or formal contracts. A striking example occurred in January 2025, when a DEA task force officer reportedly used a local Illinois detective’s Flock login to conduct dozens of “immigration violation” searches, despite Illinois law prohibiting such use of ALPR data. This informal data-sharing environment has facilitated thousands of questionable searches, including those related to immigration enforcement and even abortion tracking, sparking widespread condemnation. In response, Flock Safety has restricted access to its national lookup tool in certain states and faced a formal congressional investigation into its role in enabling invasive surveillance practices. The company’s new “Nova” platform, which integrates ALPR data with other information, has further exacerbated concerns, with reports suggesting some of its supplementary data may even originate from hacked sources, like parking meter apps.

These disparate incidents paint a cohesive picture of a society grappling with the accelerating pace of technological integration into sensitive domains. From the personal devices worn by law enforcement to the digital replication of human creativity and the pervasive reach of automated surveillance, the core challenge remains how to harness innovation responsibly while safeguarding fundamental rights and privacy in an increasingly data-driven world.