Meta tests military-grade facial recognition on smart glasses via Pentagon contractor partnership

Meta Partners with Pentagon Vendor for Smart Glasses Technology

Meta has been testing facial recognition capabilities on its Ray-Ban smart glasses using technology licensed from Rank One Computing, a Denver-based firm that earns roughly 80% of its revenue from government clients including the US military and law enforcement agencies2

2

. This marks the first known business relationship between Meta and the Pentagon vendor, raising questions about consumer tech merging with surveillance infrastructure that was previously reserved for defense and policing applications2

2

.

The partnership came to light through a WIRED investigation that uncovered a software license tying Rank One Computing to Meta's AI companion app, the same application used for the company's smart glasses products1

1

. The license authorized Meta to use military-grade facial recognition technology and liveness detection, which confirms whether a camera is viewing a live person rather than a photo or mask1

1

.

Dormant Code Shipped to 50 Million Devices

The investigation revealed that remnants of Rank One's code sat dormant inside Meta's AI app that shipped to more than 50 million phones3

3

. Meta also developed its own internal facial recognition system, internally designated as NameTag, within the same application. This feature, if activated, could have enabled Meta smart glasses to conduct real-time biometric identification of anyone in view without consent, using stored digital faceprints1

1

.

The code was never made active for users, and Meta deleted both systems one day after the news broke in early June2

2

. The Electronic Frontier Foundation's Threat Lab verified the findings and noted that while Meta reversed course following public blowback, deleting the code "does not equal a permanent change of heart"1

1

.

Military Contractor's Surveillance Capabilities

Rank One Computing is no ordinary technology vendor. Founded in 2015 and listed on Nasdaq in February 2026, the company supplies AI-powered biometric systems to the US Marshals Service, the Naval Criminal Investigative Service, and US Special Operations Command3

3

. The firm has verified prisoners for the US Marshals Service since 2021, and Special Operations Command funded research that can reportedly identify faces from up to one kilometer away2

2

.

The company's leadership reflects its deep ties to government surveillance. Its CEO previously ran the FBI division responsible for maintaining the bureau's biometric databases, while its board includes former CIA, FBI, and Pentagon officials2

2

. The license Meta acquired supported up to 10 million facial templates2

2

, demonstrating the scale at which this technology could theoretically operate.

Privacy Concerns and Accuracy Issues

The partnership intensifies privacy concerns around wearable technology becoming everyday surveillance devices. NIST testing revealed that a version of Rank One's algorithm produced false matches at different rates based on sex and country of birth, a proxy the agency uses for race, with higher error rates for women2

2

. Privacy advocacy groups have long documented how facial recognition technology disproportionately harms marginalized groups through movement tracking, misidentification of people of color, and wrongful arrests1

1

.

The US currently has almost no national rules governing facial recognition3

3

, creating a murky legal landscape around what Meta was testing. Pairing always-on camera glasses with this kind of recognition represents the scenario privacy researchers have warned about for years: identifying strangers on the street in real time without consent2

2

.

Meta's Contradictory Public Stance

Meta's experimentation contradicts its public commitments on biometric data. In late 2021, under public pressure, Meta announced plans to shut down its central facial recognition database on Facebook, which had enrolled about 600 million users globally1

1

. The company later settled a 2024 lawsuit filed by Texas over facial recognition data collection for $1.4 billion1

1

.

A Meta spokesperson told CNET the company has made no final decisions on facial recognition software for Meta Glasses, stating: "Nothing has shipped to consumers and no final decision has been made on what to do here, if anything. If we do decide to roll something out, we will take a thoughtful approach and do so with full transparency"1

1

. However, Meta declined to confirm whether it is licensing military-grade technology, when the relationship with Rank One began, or whether it continues2

2

. Rank One Computing also declined to comment on the findings1

1

.

What This Means for the Future

The ethical implications extend beyond Meta's immediate plans. The incident demonstrates how thin the line has grown between surveillance technology sold to police contractors and consumer products marketed to the general public1

1

. The technology to identify every face you look at has been licensed and tested; the remaining questions center on who ships it first and under what legal framework2

2

. Earlier this year, the New York Times reported Meta was developing software for its smart glasses to identify people using data from Facebook and Instagram, with an internal memo suggesting political tumult in the US would distract critics from the feature's release1

1

.