A new government-run identification platform in New Zealand has raised concerns about potential instances of racial bias. The Māori Reference Group at the Ministry of Social Development (MSD) acknowledges that there may be racial bias in the current iterations of the Identity Check technology. However, they express confidence that the technology will continue to improve and address these issues over time.
Facial recognition AI has been criticized globally for being less accurate when it comes to identifying individuals with darker skin tones. While some algorithms are better than others, the technology as a whole is evolving and becoming more reliable. The MSD report in September highlighted the potential risk of racial bias in the Department of Internal Affairs’ (DIA) technology, although no specific tests have been conducted to assess this risk.
Privacy Commissioner Michael Webster emphasizes the importance of agencies conducting due diligence and ensuring that the accuracy levels of the technology are robust. Webster’s office will continue to engage with officials to understand how they have addressed these concerns.
The DIA claims that recent tests with around 250 people have shown that the Identity Check system correctly matches images 90 percent of the time. However, the Māori Reference Group criticizes the DIA for not actively involving Māori and Pasifika communities in the consultation process. The group believes that robust consultation is necessary to gain trust in the technology and its usage.
Despite these concerns, MSD has proceeded with the implementation of Identity Check. It is important to note that using the system is voluntary, and beneficiaries can still choose to prove their identity by visiting a service center in person.
In conclusion, while there are concerns about potential racial bias in New Zealand’s government identification platform, there is also recognition that the technology will continue to improve. Ongoing consultation with Māori and Pasifika communities is crucial to address these concerns and ensure that the technology is used effectively and ethically.
1. Is the government identification platform mandatory?
No, the use of the Identity Check system is voluntary. Beneficiaries can still choose to verify their identity by visiting a service center.
2. How accurate is the facial recognition technology?
Recent tests have shown that the Identity Check system correctly matches images 90 percent of the time. However, there are concerns about potential racial bias, which the technology aims to address and improve upon.
3. How are Māori and Pasifika communities being consulted?
The Māori Reference Group at MSD has emphasized the need for meaningful consultation with Māori and Pasifika communities to address any issues or concerns regarding the use of the government identification platform. The group believes that their input is vital to ensuring trust and acceptance of the technology.
4. Is the privacy of individuals protected?
Privacy concerns are being taken seriously, and a privacy impact assessment is being carried out to assess any potential issues. The Privacy Commissioner’s office is actively reviewing and providing feedback on these assessments to safeguard individuals’ privacy.
(Source: Radio New Zealand)