Australia is posed to join the growing list of countries using facial recognition with its new identity system.

The technology is set to be a part of the “liveness” test, ensuring citizens are physically present human beings. Users will undertake an initial facial recognition process to establish their biometric information. Then, they will use their own mobile device to take a selfie. Finally, the picture will be analyzed by a government algorithm to confirm the person’s identity. This algorithm will electronically compare one’s personal information and the selfie to a specific government record to verify one’s identity using biometric information related to his or her facial image.

The New National Identity System

The Australian Tax Office is leading the development of this test for the myGovID platform that will allow access to sensitive information such as health and tax records. This test is a part of the government’s new national identity system set to be rolled out in the next 12 months.

At first glance, this technology has the potential to be helpful. According to a study by the U.S. National Institute of Standards and Technology (NIST), some facial recognition systems have accuracy scores as high as 99.97%. Therefore, it is possible that the technology provides Australians’ with increased security for their information and that the technology makes their lives easier. Afterall, facial recognition doesn’t require any contact, offering a quick and automatic experience.

But do the positives of this new technology outweigh the negatives?

In Australia, this question is still being answered. The Department of Home Affairs recently introduced an Identity Matching Services bill that seeks to enable a national hub for the checking of facial images on licences and other identity documents for safety purposes. The bi-partisan Senate committee, however, has recommended that the bill be rewritten to improve privacy and security. Their list of suggestions include requiring annual reporting and a restriction that automated decision making can only be used for decisions that produce favorable or neutral outcomes for the subject.

The Potential Ramifications

In essence, there are some serious concerns about facial recognition technology across government bodies.

For example, the same NIST study found that Asian, African American and Native American groups have higher false-positive rates than caucasian groups. So while the myGov technology may be increasing security for Australia’s white population, it disregards the 6 million Australians who have non-European and Indigenous backgrounds.

Furthermore, the data from facial recognition technology is often misused. One example of this is Clearview AI, a company founded by an Australian techie—Hoan Ton-That. The facial recognition app allows users to upload a photo of a person and see all the public photos of that person, along with links to where those photos appeared.

Without any federal approval or regulation, over 600 law enforcement agencies have started using Clearview over the past year. Because of the uploads from the police, Clearview now possesses a growing database of individuals who have attracted attention from law enforcement. These individuals are not informed that they have been added to this system.

This instance shows that governments are far from immune to misusing facial recognition technology. Therefore, the Australian government’s integration of this technology has the potential to become quite harmful if utilized without proper precautions.

As we look to the future, whether Australia’s integration of facial recognition technology is good or bad remains debatable. However, it has become clear that myGovID should be monitored closely, regardless. It is important that it doesn’t become the status quo to implement technology that is only beneficial to a specific subset of the public.

ABOUT THE WRITER

Sarah Raza is a Tech Innovation Fellow with a background in computer science from Stanford. She is passionate about exploring the implications of increased usage of artificial intelligence and machine learning.