Facial recognition technology: Big brother's ultimate weapon against civil liberties

Governments and law enforcement agencies are gradually implementing surveillance technologies that are more accurate, unnoticeable, cheap, pervasive, ubiquitous and searchable in real time. In public places and at work, facial recognition technology is recording and storing your image. The recent surge in popularity of social media sites has further accelerated the gathering and storing of image data.

How does civil society balance the protection of personal privacy and security in light of facial recognition technology? Are the ever-increasing security needs of the state trumping the individual's privacy and liberty?

Facebook has recently rolled out a new feature allowing members of the social networking website to be 'tagged' in photos (uploaded by others). Facebook's 'tag suggestions' system uses facial recognition software to identify individuals as photos are uploaded and 'friends' are encouraged to put names to the faces. It is increasingly coming under criticism by those worried about privacy.

The feature allows users to identify their 'friends' in photos without their permission. While it is possible to disable tagging, the default setting enables the feature automatically. Furthermore, if users don't want their name attached to an image they need to 'untag' themselves after the photo has been uploaded.

Anecdotal evidence suggests that the vast majority of the 600 million members accept the default settings. With approximately two million photographs uploaded to Facebook each day, every new image is now processed by automatic face recognition software. But this technology it is not just restricted to the social media.

The Federal Bureau of Investigation (FBI) has already initiated the Next Generation Identification (NGI) program. This program will further advance the FBI's biometric identification services, providing an incremental replacement of current Integrated Automated Fingerprint Identification System (IAFIS) capabilities. According to the FBI: "The NGI system will offer state-of-the-art biometric identification services."

In the United States, 18 000 law enforcement agencies already contribute over 200 000 fingerprints and DNA samples to the FBI's databases everyday. Under NGI an iris (eye) print database is being added to the existing fingerprint and DNA databases. Technology now allows fingerprints to be amplified with friction prints of other ridges, identifies palm-prints and footprints. The introduction of voiceprints is currently being evaluated. Any conceived measure, according to the FBI, that can feasibly increase public safety is being considered.

For law enforcement officials, facial recognition would be the "killer application of biometrics". In 2009, a senior FBI technologist claimed that the FBI would "dearly love to be able to use facial recognition in its fight against crime." At the time that this statement was made, the algorithms didn't exist to deliver the highly reliable verification required. Facial recognition software use feature-extraction algorithms to find patterns in skin texture and in the curves of the eye sockets, chin and nose. Nevertheless, it is not an exact science, and computers can get it wrong. The FBI had been considering facial recognition technology since 1963 but since then had not invested enough in developing the technology. Facebook and others have now provided it as affordable technology.

The British government has announced that it would like to check the identity of everyone entering and leaving the U.K. by 2013 with facial recognition technology. The U.K. Border Agency already collects biometric information. They require visa applicants to provide scanned fingerprints and have a digital photograph taken when they make a visa application. In Australia, the National Security Science and Technology Branch (NSST) is a division within the Department of Prime Minister and Cabinet (DPC). The NSST "provides a national focus for science and innovation aimed at enhancing Australia's national security." Australia is introducing a biometric based visa system for certain non-citizens.

Since 2010, the Department of Immigration and Citizenship (DIAC) collects biometrics (fingerprints and facial images) from visa applicants in selected overseas locations (they seem reluctant to name which ones). According to DPC: "This data will be matched against DIAC's biometric database and against the biometric databases of international partners. This initiative will reduce the risk of terrorists, criminals and other persons of concern entering Australia undetected".

Sounding similar to the U.S program, Australia's Next Generation Border Security is described as the use of "advanced data analysis and risk profiling to better identify visa applicants who may present national security risks to our intelligence agencies." It is expected to cost the Australian government $69 million over four years (2010-2014).

European Union data protection regulators have said they would investigate Facebook over its use of facial recognition software and its 'tag suggestion' to name people in photos without their knowledge and uploaded without their permission. It is thought the measure is a possible rule violation under Article 29 of the E.U. Data Protection Working Party. Google, Microsoft and Yahoo have already been pushed by European data-protection officials to limit the amount of time they store online users' search records.

But how reliable is face recognition technology? Can computers really identify a human face from an image caught on a surveillance camera? With increasing uses of public camera's at airports, in streets and in public buildings all hooked into computer networks unknown to the individual being filmed, could computers be trying to match your face to databases of criminals? There are two issues here: What if the technology gets it wrong and you are tagged as someone you are not? And does the government's increasing use of such technology reinforce fears of a Big Brother state?

Technologist Tim O'Reilly's view is that face recognition is here to stay. The question is "whether to pretend that it doesn't exist, and leave its use to government agencies, repressive regimes, marketing data mining firms, insurance companies, and other monolithic entities, or whether to come to grips with it as a society by making it commonplace and useful, figuring out the downsides, and regulating those downsides".

While Facebook has been criticised for its photo tag feature, the larger concern is the use of facial recognition software in public spaces and how it is being used. The NSW government, or more specifically the NSW police, is running facial recognition software against CCTV footage. The NSW RTA is also using the same software and same footage to link images to the RTA database of driver's license photos. The Australian federal police have access to both agencies. The federal body CrimTrac has asked NSW for its database so it can be mined nationally by police using the facial recognition information contained in it.

In 1987 Australian's strongly rejected an Australia Card because of privacy issues, with the unregulated use of facial recognition technology and computer networks, it appears society has come a long way in its willingness to give up civil liberties in return for apparent safety and protection.