Despite not making the full policy public, the US Department of Justice (DOJ) has revealed insights into its interim policy concerning facial recognition technology (FRT). The testimony submitted to the US Commission on Civil Rights highlights key aspects of the policy announced in December, emphasising its adherence to protecting First Amendment activities. The policy aims to prevent unlawful use of FRT, establish guidelines for compliant use, and address various aspects, including privacy protection, civil rights, and accuracy.

Ethical considerations are integral to the interim policy, with measures in place to prevent discriminatory use of facial recognition and ensure accountability for its deployment. However, complexities arise due to evolving AI regulations and the proliferation of biometric algorithms, leading to stipulations that FRT systems must comply with DOJ policies on AI and that FRT results alone cannot serve as sole proof of identity.

The testimony acknowledged civil rights concerns, recognising the potential for bias in algorithms and the misuse of FRT, including unlawful surveillance. Nonetheless, the DOJ emphasises the benefits of FRT in enhancing public safety, citing its role in identifying missing persons, combating human trafficking, and aiding in criminal investigations. According to the DOJ, the key lies in harnessing FRT's potential while implementing effective safeguards to mitigate potential harm.

Why does it matter?

In a related development, the US government has recently published new guidelines that require all federal agencies to appoint senior leaders as chief AI officers to oversee the use of AI systems. According to the guidelines, agencies must establish AI governance boards to coordinate usage and submit annual reports detailing AI systems, associated risks, and mitigation strategies. As a result, the US Department of Justice appointed Jonathan Mayer, an assistant professor specialising in national security, consumer privacy, and criminal procedure at Princeton University, as its first chief AI officer.