Meta Tested Rank One Computing Software

Meta licensed Rank One Computing’s software for testing with its smart glasses. This isn't just a gadget. But it's a direct link between consumer wearables and high-grade military surveillance tools, and that raises serious questions about privacy and the ethics of facial recognition technology in everyday life. The arrangement, documented in an active software license, shows Meta integrated the company’s face-recognition capabilities into a test version of the Meta AI companion app. So this software powers the Ray-Ban and Oakley smart glasses, and it marks a clear step in Meta’s exploration of facial identification features for everyday consumer hardware, and they're not backing away from that.

This Denver-based developer went public on the Nasdaq in February. It generates about 80 percent of its revenue from government contracts, and its technologies are actively used by the United States military, federal law enforcement agencies, and local police departments. But the discovery of the license highlights how the division between defense-grade surveillance tools and mass-market consumer electronics is rapidly disappearing, as identical underlying algorithms begin to serve both sectors. It's a vanishing line.

The Mechanics of the Software Integration

Meta's license authorizes the face-recognition engine and specialized liveness detection. It's a verification system that ensures the camera is capturing a living, breathing person instead of a static photograph or a mask, and the active agreement supports a database of up to 10 million distinct facial templates. But that's a lot of data.

Software code reviewed by WIRED revealed that integration routines for loading the license and initializing the software remained inside a version of the Meta AI app shipped to millions of consumers. But they were dormant. These routines existed alongside Meta's own internal face-recognition systems, though none of these capabilities were ever made active for the public.

Meta removed the code entirely on June 5. This action occurred a day after WIRED revealed that the company had built an unreleased, dormant face-recognition system internally named NameTag into the companion app, which has been downloaded to more than 50 million smartphones. Meta declined to answer questions regarding why it licensed the software, when the partnership began, or if the relationship remains active. The software developer also declined to comment.

From the Pentagon to Consumer Pockets

Military tech routinely migrates. Joseph Jerome, a former Meta Reality Labs policy official, pointed out the historical path of these developments.

"There's a long history of military technologies becoming consumer products. That's the story of the internet, in many ways."

Rank One Computing's technology has a deep footprint in defense and law enforcement. It's essential. Founded in 2015 by engineers who previously built face-recognition systems at the nonprofit research institute Noblis, the company's leadership is heavily populated by former national security officials. But B. Scott Swann, the CEO, used to run the FBI division that managed the bureau's biometric databases. The board includes a former CIA deputy director for science and technology, a former head of the FBI's science and technology branch, and a former Pentagon official who established a multibillion-dollar special-capabilities office.

Active Government Deployments

• The US Marshals Service has used a biometric identification kit powered by the company's technology since 2021 to verify prisoners' identities without fingerprinting during transport.
• The Naval Criminal Investigative Service purchased the company's specialized video tool, ROC Watch, for its police operations.
• The company developed long-range face-recognition capabilities for US Special Operations Command, designed to identify faces from a distance of up to one kilometer.
• Dozens of schools in West Virginia have used the software at building entrances to screen visitors against the state's sex-offender registry.

Commercial Software Bundling

The company's algorithms are embedded in third-party platforms used by local police. DataWorks Plus, a South Carolina firm, bundles the technology into its own products. But it doesn't stop there. LexisNexis also includes the algorithm inside its Lumen platform, which lets police officers run instant searches against state, regional, and national FBI database images.

The Demographic Accuracy Gap

It doesn't perform equally across all groups. Testing by the National Institute of Standards and Technology found that a version of the algorithm produced false matches at highly variable rates depending on a person's sex and country of birth. So the federal institute uses country of birth as a proxy for race during its evaluations.

Error rates were lowest for individuals born in Eastern Europe. But the algorithm produced higher error rates for women than for men, and that discrepancy raises serious questions about whether we can really trust these systems when they're deployed at a massive consumer scale.

A Regulatory Void in the United States

Regulatory oversight of biometric tracking is highly fragmented across the country. It's a patchwork mess. While some states require law enforcement officers to obtain a warrant before accessing facial recognition databases, a cohesive national framework does not exist, and the pace of technology continues to outrun legislative efforts every year. But more states are incorporating biometric protections into broader consumer-privacy laws. We can't keep up.

Eric Null, director of the Privacy and Data Project at the Center for Democracy and Technology, highlighted the growing corporate demand for these systems.

"But consumer-facing companies clearly crave access to high-powered facial recognition technology. And without proper checks, the risks of this tech becoming a common consumer product are large and largely unbounded."

They're testing it quietly. But the integration of defense-grade identification systems into consumer wearable hardware remains a highly sensitive frontier that consumer electronics companies continue to develop with great caution. That's the reality. A major social media firm's quiet testing of Rank One Computing software emphasizes just how close these powerful surveillance capabilities are to entering the consumer mainstream.