Artist's campaign targets biometric surveillance
Artist and researcher Adam Harvey has set out to raise awareness about the increasing pervasiveness of biometric tracking on the web and in everyday life.
Artist and researcher Adam Harvey has set out to raise awareness about the increasing pervasiveness of biometric tracking on the web and in everyday life.
"Today's selfie is聽tomorrow's聽biometric portrait," read聽the bright red text of a poster聽plastered to the New Museum鈥檚 window in New York City鈥檚 Bowery neighborhood.
It's a simple, arresting message that artist,聽researcher,聽and privacy activist Adam Harvey hopes will compel聽anyone who sees it to reconsider how they use technology, especially as internet companies and government agencies deploy increasingly sophisticated biometric software to track, surveil, or identify individuals.
"If people knew that a company was quantifying and commercializing their age, race, gender, mood, clothing, the amount of lipstick they wear, the amount their midriff is exposed, and the expression on their face, would they act differently? I think so," says Mr. Harvey. "The problem is we can't see how we're being seen. And as a result we can't know how to act accordingly."
Based in Berlin, Harvey is among a growing number of artists who are gaining attention for their focus on digital issues in the wake of the Edward Snowden revelations of widespread National Security Agency surveillance.
While artists such as Laura Poitras, who directed the documentary about Mr. Snowden "CitizenFour," and British graffiti artist Banksy have focused on mass surveillance and US drone policies, Harvey鈥檚 work attempts to raise awareness about the use of biometric technologies to record and track individuals鈥 unique characteristics such as facial features, speech patterns, and even their manner of walking.聽
Pieces in Harvey鈥檚 "Think Privacy" series that includes the selfie poster聽have聽appeared聽online as well as聽in galleries from Hong Kong to Berlin.聽And as聽he expands the collection this month, biometric聽surveillance聽is advancing rapidly.聽
In fact,聽computers are now more than聽98 percent accurate when it comes to聽recognizing faces,聽according to a Chinese University of Hong Kong research paper. And, machines appear to be getting smarter at identifying specific people due to the massive amounts of photos uploaded to the web.
But even though聽commercial stores,聽law enforcement, and intelligence agencies justify using聽this technology to detect shoplifters聽or to suss out criminals and terrorists, Harvey says that most people remain uncomfortable with biometric software. Specifically,聽he pointed to a survey from research聽firm First Insight that found that 75 percent of respondents wouldn't shop in a store that used facial recognition technology.聽
Some consumers have even聽started pushing back against companies that use biometrics.聽Both Google and Facebook are facing a聽lawsuit over the collection and storage of their users鈥 facial images. The case claims that both tech giants unlawfully collected the information in violation of Illinois鈥檚 Biometric Information Privacy Act. The outcome of the case could聽set a precedent for how the technology evolves in the US. In Europe, regulators already have strict rules that limit facial tagging on the internet.聽
Many tech advocacy groups in the US also want more transparency around the use of facial recognition technology. In a letter to the Justice Department, a group of 45 civil liberty and privacy groups opposed an FBI plan to聽exempt its biometric database from Privacy Act provisions.
But biometric technology is hardly new. Tools first聽emerged the 1960s when a Japanese team created software to detect the outline of someone鈥檚 head. By the 1970s, computers could pinpoint eyes, noses, and mouths. For the first time in the 1990s, the Department of Defense created a large database to spot combatants at a distance, and by 2001 the standard聽"Viola Jones" detector, which is still in use, was created.聽
Yet it wasn't until 2010 that biometrics emerged into the mainstream, and these formerly military technologies were for the first time applied to a slew of celebrity look-alike websites, says Harvey. Many聽Internet聽users freely sacrificed their information to be paired with their star doppelganger on Facebook.
"We were so busy having fun that we didn鈥檛 realize we were training the largest facial recognition device in the world," says Harvey.
Since then, researchers have created websites such as聽MegaFace that include聽vast archives of photos gleaned from the internet. While this data is largely collected from commercial sources, it's聽also gathered for governmental purposes, says Harvey. Starting in 2012, programs such as Janus from the government's Intelligence Advance Research Projects Activity (IARPA) agency have turned to social media to gather聽biometric data.
"The IARPA Janus program was developed to remotely identify an enemy of the state while Facebook's goal is to know everything about you in order to sell 鈥榓dvertisements,' " says Harvey. "But is it still an 'advertisement' when it's also a psychological taunt driven by statistical models of your consumer behavior? I'd argue that primary use in both cases is social control."
Such information is also being "power-grabbed" from the Internet without regards to ethics or user consent, says Harvey, and used for academic purposes, such as a one recent聽Twitter facial analysis聽that reveals the demographics of presidential campaign followers. Google recently backed聽a research paper on how to use clothing聽鈥撀爏tudied from public but copyrighted online images 鈥 as metadata.
Havey is optimistic that as the public becomes aware of the pervasiveness of biometric tracking they will advocate change. For instance, he said, a decade ago it was difficult to find easy-to-use software that could encrypt sensitive communications. Now, that technology is on every iPhone.
People need to have a better way of avoiding biometric detection than just 鈥渙pting out of dubious tracking technologies every day,鈥 he said. 鈥淵ou can鈥檛 make a good decision if you are only told the benefits and the costs are hidden.鈥