Australia's AI Face Scanning Trial: A Troubling Development
The recent news of Australia's first live AI face scanning trial has sparked intense debate and raised serious concerns among legal experts and academics. This trial, conducted by Western Australia Police, has been criticized as a "mass surveillance exercise" with potential implications for privacy, accuracy, and even racial profiling.
The Trial and Its Impact
The trial, which began in June, involved marked vans equipped with cameras scanning the faces of individuals in real-time and matching them against a watchlist of approximately 4,000 people. In just the first week, an astonishing 130,000 faces were scanned, leading to 33 alerts and 18 arrests in the state's capital. This raises immediate questions about the scope and intensity of this surveillance.
Ethical and Legal Concerns
Legal experts argue that the use of such technology in public spaces, especially when accessing sensitive personal data on a large scale, is fraught with ethical and legal pitfalls. Greg Barns SC, a spokesperson for the Australian Lawyers Alliance, warns that this technology could be misused to surveil and monitor individuals engaged in lawful activities, such as attending protests or simply gathering with others in public.
"It's surveillance without a warrant, which is normally required to surveil a person," Barns points out. This lack of oversight and potential for abuse is a major cause for concern.
Privacy Laws and Misidentification
What makes Australia particularly vulnerable is its weaker privacy laws compared to other countries. As Barns explains, if an individual's privacy is breached, the law may not offer sufficient protection or the ability to seek damages, as is the case in the UK.
This is especially worrying when considering the potential for misidentification. If innocent people are wrongly accused of criminal offenses, Australian laws may not provide them with adequate legal recourse.
Lack of Transparency and 'Function Creep'
The trial has also been criticized for its lack of transparency. Anna Zenz, a legal academic at the University of Western Australia, finds it striking that WA Police did not formally consult the state's privacy watchdog before initiating the trial. This lack of consultation raises questions about accountability and the potential for abuse.
Zenz echoes privacy concerns, arguing that these technologies not only invade personal privacy but also erode our collective privacy. The constant feeling of being watched in public spaces can lead to changes in behavior and a decline in trust levels. Moreover, these technologies have a tendency to expand beyond their initial intended use, a phenomenon Zenz describes as "Function Creep." This could lead to a trial becoming permanent, overt use becoming covert, and a change in the context of use.
Despite WA Police's assurance that no data is recorded or kept, Zenz highlights valid concerns about data security and hacking. Given the sensitive nature of the data, cybersecurity risks are significant and cannot be ignored.
Bias and Racial Profiling
Experts also express concerns about the AI model's lack of training on the diversity of the Australian population. Evidence suggests that minority groups are at a higher risk of misidentification, potentially exacerbating existing biases within the policing and criminal justice systems. This technology could further target and overpolice already marginalized communities, such as indigenous Australians, as seen in countries like the USA and the UK where similar technologies have been used to profile black people.
The Benefits vs. Concerns Debate
WA Police Commissioner Col Blanch argues that the system is a policing tool, not a general surveillance network, and is about "protecting people, not watching people." However, the concerns raised by experts and academics suggest that the potential for abuse and the erosion of privacy rights may outweigh the benefits of this technology.
The Way Forward
Experts are calling for greater transparency from WA Police and independent oversight of the technology's use to protect public rights. As Jessamy Perriam, a digital sociologist at the Australian National University, notes, while there is a desire for public places to feel safe, it must be achieved in a way that respects people's privacy and boundaries. This trial serves as a reminder of the delicate balance between security and civil liberties, and the need for careful consideration and oversight when implementing such technologies.
