How To Defeat Facial Recognition Software

You probably don’t know it (or maybe you do) but your face is likely in a massive police database. News broke that law enforcement has over 117 million faces stored. Here’s what you can do to get around this.

Further Reading

Are You Infected with Malware? Here’s How To Find Out

British Companies Sell StingRays To Oppressive Governments

Faces Everywhere

The Center for Privacy & Technology at the Georgetown University law school released a huge report today. It gives details on a facial recognition database that law enforcement such as the FBI use. It contains roughly half of American adult faces – which is about 117 million faces. And this database of biometric data doesn’t have a lot of oversight, which means it’s easy for the government to abuse it.

The report mentions the dangers to privacy, free speech and protections against unreasonable search and seizure that this data enables. Right now, the report finds that about a quarter of all local and state police departments can access this database (dataface?). Additionally, law enforcement in more than half of the United States can search this database and compare the faces to ID photos like your drivers’ license.

Alvaro Bedoya, executive director of the Center for Privacy & Technology, told Wired,

“Face recognition technology lets the police identify you from far away and in secret without ever talking to you…Unless you’ve been arrested, the chances are you’re not in a criminal DNA database either, yet by standing or a driver’s license photo at least 117 million adults have been enrolled in a face recognition network searched by the police or the FBI.”

So far, no state has passed legislation to give better parameters of how this facial recognition should be used in police investigations. A small number of departments around the US have imposed voluntary restrictions on searches. They require reasonable suspicion or necessitate that the data only be used in serious crime cases.

At the same time, no department has standards for testing the accuracy of their systems. No police staff has formal training to recognize face matches either. You’d think that this would be an innate skill, but it actually requires special training. There are also concerns of racial bias.


Many law enforcement agencies have argued that these biometric tools reduce racial profiling. After all, shouldn’t a computer algorithm be free from human prejudice? Not exactly. If a human programs/creates these algorithms, sometimes prejudice can slip in unnoticed. It depends on the data sets that researchers use to train machine learning systems. Sometimes the systems are better at identifying people of a particular race than others.

Even more scary is the potential for dystopian governments to perform real-time facial recognition. Every time you pass a surveillance camera in public (or even in private) the software can instantly recognize you, and track your whereabouts. Researchers found that five police departments in big cities like Los Angeles and Chicago already use live recognition.

In the report, a coalition of 40 civil rights and civil liberties groups started an initiative on Tuesday. They’re asking the Department of Justice’s Civil Rights Division to more closely study the current use of facial recognition technology. Neema Singh Guliani, legislative counsel for the ACLU said,

“Police are free to identify and potentially track anyone even if they have no evidence that that person has done anything wrong…We don’t expect that the police can identify us when we are walking into a mosque, attending an AA meeting, or when we’re seeking help at a domestic violence center.”

What You Can Do

Fortunately, there are methods to bypass facial surveillance. Dan Moren of Popular Science wanted to fool his banking app. The app had a biometric feature where it required a live feed of you blinking to make sure it was actually you. But he was able to fool the app. He first printed out a photograph of his own face and cut holes for his eyes. But the app didn’t recognize the portrait.

Then, he recorded a video of his whole face, where he blinked at the camera. The app recognized that and he logged in successfully. A hacker could easily take control of your laptop’s camera to record a short video of you and use that to log into the banking app to steal your money.

You can trick facial recognition in other ways too, using special hairstyles and makeup. The website CV Dazzle offers different camouflage methods, based on six style tips:

  • Makeup: Avoid enhancers that amplify facial features. Instead, use makeup that contrasts with your skin tone in unusual tones and directions. Examples: light colors on dark skin, dark colors on light skin.
  • Nose Bridge: Partially cover the nose bridge area. This is where the nose, eyes and forehead intersect to form a key facial feature.
  • Eyes: Partially obscure one of the ocular regions, such as position and darkness of eyes.
  • Masks: Don’t wear masks because they can be illegal in certain cities. What you want to do instead is change the contrast, tonal gradients, and spatial relationship of dark and light areas on your face using hair makeup and/or special fashion accessories.
  • Head: Certain research shows that obscuring the elliptical shape of your head can improve your ability to evade facial detection.
  • Asymmetry: Facial recognition software expects to find symmetry between the left and right sides of your face. Break this up by developing an asymmetrical look.
Images from CV Dazzle

Images from CV Dazzle

“From all appearances, deception has always been critical to daily survival – for human and non-human creatures alike – and, judging by its current ubiquity, there is no end in immediate sight.”

– Roy Behrens, Camoupedia


Can you see yourself rocking one of these looks? The first thing I thought was these guys look like something out of a Dystopian sci-fi world. Let us know what you think in the comments below.