Getting your Trinity Audio player ready...
|
New Zealand is becoming increasingly diverse as a multicultural hub, with various ethnicities contributing to creating new cultural identities by bringing their unique heritage. Managing the data about these diverse populations poses a challenge for the government as it strives to accurately capture the richness and complexity of the cultural landscape shaped by the amalgamation of different races and identities within the country, creating an intricate racial bias.
In light of this, the New Zealand government is set to roll out a facial recognition technology to confront this issue in the digital era in the digital era. Despite concerns about potential racial bias, this cutting-edge technology, developed over four years by the Department of Internal Affairs (DIA), aims to enhance identity verification processes, particularly for beneficiaries and government services.
The new system utilises facial recognition technology, which promises to transform how individuals interact with government services, offering seamless and secure identity verification. Users can match a live image taken on their phones with their driver’s license or passport photo stored in government databases. The Ministry of Social Development (MSD) is set to implement this identity verification system today, marking its first deployment phase.
As highlighted in recent reports, the concerns surrounding potential racial bias in facial recognition technology have spurred a closer examination of the implications within New Zealand’s distinct demographic landscape. The unique population mix, encompassing diverse ethnicities and cultures, emphasises the need for a comprehensive understanding of how this identity verification system functions across different groups.
The Department of Internal Affairs (DIA), the driving force behind this identity verification system, has affirmed its commitment to addressing concerns related to racial bias. Despite not conducting specific tests on ethnic groups within the country, the DIA’s recent difficulties, involving 250 individuals, yielded a commendable 90% accuracy rate. This success rate is encouraging, suggesting that the technology is well-calibrated to the nuances of the New Zealand context.
Despite the potential unknowns surrounding racial bias, the MSD sees the risk as acceptable, with ongoing efforts to consult ethnic groups and improve the technology continually. The MSD report acknowledges the concerns but emphasises that engagement with Pasifika groups has occurred, and discussions with Māori groups are anticipated in the next quarter.
While acknowledging the potential risks, the government emphasises the positive impact of this identity verification system in simplifying access to benefits and securing online services. Users, primarily beneficiaries, are given five attempts before being locked out for three days. Furthermore, I this identity verification system is designed not only for user convenience but also to detect and investigate fraud, although fraud investigation purposes will not be implemented in the initial phase.
However, the rollout faces criticism from AI researcher, Joy Liddicoat, who argues that launching a system without addressing known risks over a four-year project is inadequate. Despite concerns, the government is determined to make this identity verification system a go-to online verification technology. It plans to expand its use across various services and e-business transactions. The government’s commitment to acknowledging the untested risk and engaging with ethnic groups after the system goes live demonstrates a proactive approach to address concerns transparently.
“As this identity verification system becomes integral to the digital identity landscape, ongoing collaboration, refinement, and communication will ensure the technology’s success and inclusivity,” said the government’s representative.