A US federal watchdog has called for clearer guidelines for public housing agencies on the use of facial recognition technologies.
The General Accounting Office made the recommendation as part of a review of the use of AI in the rental housing sector.
It said that “property technology” was widely used in the housing market, whether for “advertising, touring, leasing, and financial management of rental housing.”
These tools could incorporate algorithms and artificial intelligence, the GAO said, and could also be used for tenant screening, setting rents, or for security and controlling building access.
But, it added, “these tools also may pose risks related to transparency, discriminatory outcomes, and privacy.”
Potential renters and owners might struggle to understand or explain algorithmically driven decisions.
Meanwhile, “Facial recognition systems also might misidentify individuals from certain demographic groups, and property owners might use surveillance information without renter consent.”
The study covered four federal agencies, as well as 10 public housing agencies. The GAO noted that the federal agencies had taken action to “combat alleged misleading and discriminatory advertising on rental platforms” and taken enforcement actions against “tenant screening companies for using inaccurate or outdated data.”
But all the housing agencies said they could benefit from “additional direction on use of facial recognition technology.”
The current guidance from the Department of Housing and Urban Development was “high level and does not provide specific direction on key operational issues, such as managing privacy risks or sharing data with law enforcement.”
The GAO’s full report added that “Advocacy groups we interviewed expressed concerns about facial recognition technology’s higher error rates for identifying and verifying individuals from certain demographics—particularly Black women.” This could result in “frequent access denials” for some individuals.
There were also challenges around ensuring “consent” around these systems, given they rely on capturing unique biometric data.
The report highlighted other risks, such as whether “rent setting” software could leave to reduced bargaining power and increases in rental costs overall.
Quite apart from the privacy implications, facial recognition technology has long been dogged by questions over its accuracy, particularly when it comes to specific groups.
But that hasn’t stopped governments, and businesses rolling out the technology. In the UK, the Metropolitan Police is scaling up its use of the technology to identify wanted suspects in crowds.
