US Army contracted two firms for $4.5M to create a facial recognition system that analyzes infrared images to identify people at a distance of 10 to 500 meters
Dave Gershgorn / OneZero : Tweets: @ozm , @offdutytiger , @ozm , @6eorge6 , @ozm , @a_bacci , and @sophielgilbert . Thanks: @davegershgorn Tweets: OneZero / @ozm : NEW: The U.S. military is developing facial recognition technology that reads heat emitted by faces. The tech would work in the dark and across long distances, according to contracts posted on a federal spending database. http://read.medium.com/9IYcoGi Off Duty Tiger / @offdutytiger : Wearing more caps, masks, and layers, here I come. https://twitter.com/... OneZero / @ozm : Facial recognition is already employed by the military, which uses the technology to identify individuals on the battlefield. But existing tech typically relies on images generated by standard cameras, such as those found in iPhone or CCTV networks. http://read.medium.com/9IYcoGi G.K.R. Spookyhaus / @6eorge6 : Starting a line of balaclavas with built-in heat packs. “Perfect for cold weather,” I exclaim loudly toward my smart speaker. https://twitter.com/... OneZero / @ozm : Relatively little is known about how the military uses facial recognition, but continued investment in the technology suggests that the military sees it as a priority. http://read.medium.com/vJtsATC Anna Bacciarelli / @a_bacci : Facial recognition tech shouldn't be thought of in isolation, with individuals identified only via images. The reality is that integration of FR with gait/voice/vein/heat analysis is the logical next step in a new age of pervasive networked surveillance. https://onezero.medium.com/... Dr. Sophie Gilbert / @sophielgilbert : First though: wow I wonder if we could do this with wildlife and game cams?! Second thought: OMG creepy for humans! https://twitter.com/... Thanks: @davegershgorn
Context & Ripple Effects
The Army's $4.5M award to two firms extends a biometric program the outlet already mapped: the Automated Biometric Information System, a 7.4M-identity database underpinning military intelligence operations. Visible-light facial recognition, as the FT's survey of government-requestable face datasets showed, is already trained on broadly available imagery — the new contracts target its two hard limits: darkness and distance, with identification specified at 10 to 500 meters via heat signatures.
The move lands amid a wave of scrutiny captured in The California Sunday Magazine's primer on who uses the technology and how to avoid it, and after the NYT's ~$60 demonstration that off-the-shelf cloud recognition plus city cameras could identify strangers without their knowledge. Thermal recognition attacks the evasion layer that primer documents — masks, darkness, distance.
First-order effects
- The two contracted firms now hold a funded path to solve thermal face matching at standoff range — a capability gap in the Army's existing biometric stack, which identifies people from conventional imagery fed into ABIS.
Second-order effects
- Any working system plugs directly into the Army's 7.4M-identity biometric database, converting heat signatures collected in the field into matches against existing identities — making the database, not the sensor, the strategic asset.
Third-order effects
- If thermal recognition matures, the practical countermeasures documented for civilian facial recognition (covering the face, avoiding light, keeping distance) stop working, shifting the debate from camera placement to whether identification at 500 meters in the dark should be permissible at all — and the later FBI and Pentagon research that fed the Horus search tool suggests military identification programs have a record of migrating toward domestic use.
The trend: Military biometric identification is moving from controlled, close-range capture toward passive standoff recognition in any conditions, with shared identity databases turning each new sensing modality into a force multiplier.