Nowhere to run: drones, facial recognition, soft biometrics and threat assessments
By End the Lie
It is the stuff of science fiction: a drone flies hundreds of feet overhead, rapidly snapping images and collating them into a 3D model of your face, verifying your identity, recording your social interactions and even creating threat assessments of yourself and those you associate with.
Unfortunately, this is not science fiction. This is real technology being developed as you read this under several military contracts, all paid for by the American taxpayer adding on to the black hole of debt which continues to grow unabated thanks to unnecessary spending like this.
What is worse is that like most military technology, we can expect this new paradigm of war to bleed into domestic police activities and so-called homeland security operations.
The major issue brought up here, aside from the glaring privacy concerns, is that this technology can and will be used to treat suspects as guilty before proven innocent.
If you are not a fan of the government or prefer to not have your privacy violated, you can bet technology like the Department of Homeland Security’s Future Attribute Screening Technology (FAST) will set off some alarms.
Then, based on your alleged malicious intent, you will be treated as if they have any evidence that you pose a threat. How exactly they will treat this evidence is anyone’s guess but if the past is any indicator, it won’t be pretty.
This is precisely the idea behind contracts granted to the likes of Charles River Analytics which is developing a technology called Adversary Behavior Acquisition, Collection, Understanding, and Summarization (ABACUS).
ABACUS applies “a human behavior modeling and simulation engine” to data already collected by drones, intercepted phone calls and informants. It then produces “intent-based threat assessments of individuals and groups” as highlighted by Wired‘s Danger Room.
Producing a so-called threat assessment would be ineffective without knowing exactly who the subject is, and that’s where Progeny Systems Corporation comes in.
According to their official website, Progeny has fulfilled contracts since 1995 for the United States Army, United States Navy, United States Air Force, National Institutes of Health, Defense Advanced Research Projects Agency (DARPA) and Fortune 500 customers.
Among these is a contract recently issued for “Long Range, Non-cooperative, Biometric Tagging, Tracking and Location”.
This technology is “Non-cooperative” in that it does not require the subject’s knowledge or consent like traditional biometrics which are considered “cooperative” and thus “generally not applicable to the more difficult problem of ‘real-world’ recognition.”
Progeny’s technology is designed to be deployed on an Unmanned Aerial Vehicle (UAV) platform using existing UAVs and sensors. They have developed algorithms that can perform tagging, tracking, and locating (TTL) tasks on unwitting individuals with pre-existing video sources, just like another contract awardee Intelligent Automation, Inc.
Both Progeny and IAI have developed technologies that use existing UAV payloads to identify and track non-cooperative targets in real time in both urban and rural environments.
Now UAVs will be able to pick up lost targets near instantly, detect potential targets and create 3D images of their faces for future tracking, and create a threat assessment.
To remove the previous temporal constraints that made much of facial recognition too cumbersome to use in the field, Progeny also developed technology that employs “soft biometrics” to filter through crowds and “port and border monitoring scenarios to categorize individuals at a great distance.”
Without a doubt, “soft biometrics” is a nice way of saying computerized racial profiling. The algorithm cuts down on the time taken to identify individuals by sorting them “based on gender, skin color, height, weight, anatomical proportions, geometrical facial features”.
It is not farfetched to speculate that when they speak of “port and border monitoring scenarios” they would “categorize individuals at a great distance” based on ethnicity. The fact that they are promoting this as a potential law enforcement tool just emphasizes those concerns about systematic racial stereotyping.
After the Progeny system captures an image with a mere 50 pixels between the subject’s eyes, the process of building a 3D model begins. After the model is created, further images are used to create a more accurate facial model which can then be used so only about 20 pixels of image is required to identify the individual.
Crowds? No problem. Only a sliver of the suspect’s face is in the picture? No problem. No evidence to detain them? No problem, just run a threat assessment and then claim he is harboring malicious intent and the algorithm proves it.
Does anyone else see how these could create some serious problems if it were implemented at home?
Even saying “if” is a bit naïve seeing as shockingly similar threat assessment technology was being tested months ago by the Department of Homeland Security.
Seeing as police departments and DHS are increasingly utilizing drones domestically (even so-called “micro drones”), all it would take is the application of the new software (remember, it all uses existing sensors and UAVs) and we have a full-blown pre-crime police state in America.
While all of this technology has the real potential of cutting down on civilians slaughtered by drones in attacks on “suspected militants” it also has the much more grim potential of being used to wrongly accuse and/or detain innocent Americans while violating the privacy rights of each and every citizen.
This is not one of those situations where the logic of “if you have nothing to hide then why worry?” will work, if indeed such logic ever does. If we allow this type of science fiction absurdity to be forwarded on our own dime, we can expect what shred of liberty we have left under the PATRIOT Act to be eradicated.
It is a sad reality that the FBI already does not need any probable cause whatsoever to conduct surveillance on you, rifle through your trash, tap your phones, etc. we can expect this to be much worse if they can produce an “objective” threat assessment that says you’re malicious. What little need for evidence still existed in the so-called justice system will be gone in the blink of an eye.
I know these threats seem abstract or unreal but they are at our doorstep and if we sit by in silence we can expect more of our non-existent public funds to be squandered on projects that are everything that America is not supposed to be.