Gartner to biometrics firms: Do better with facial recognition. Somehow
The facial recognition community continues to speak in circles when ethics is the topic.
Notable business-to-business tech analyst Gartner is the latest with indistinct advice that amounts to “Please use better judgment.”
A new post by Gartner analyst Frank Buytendijk advises, “The appropriate use of facial recognition technology depends on the prevailing culture, ethics, legislation and practices.”
People in public might as well be fish in a barrel analyzed for profit by machine vision. Vendors leave questions of ethics in the hands of buyers. Buyers consult lawyers who wait on judges and politicians.
Facial recognition researchers worldwide express concern, and are asking pointed questions about their own practices. For example, 25 percent of researchers questioned said it is ethical to do biometric research on vulnerable populations as long as consent is given. Yet 71 percent said such studies are or could be unethical.
Some, at least, seem ready to even give up some research contracts. Forty percent of researchers questioned by the science journal Nature said public real-time face biometrics should be banned.
There is some self-reflection in the community. A poll during the Biometrics Institute’s online congress showed that virtually everyone feels the sector needs standardized testing to gain the public’s trust.
The Government Accounting Office, which advises members of the U.S. Congress, has come further than most in terms of concrete actions to investigate.
In a report published in August, the GAO found seven federal laws that could cover biometrics. Investigating the laws would be a good step for the community. For its part, the office has recommended that Congress specifically bolster privacy protections. The SIA also includes privacy by design among the ten core principles it offers up for “responsible and ethical” facial recognition use.
In the absence of coherent protocols, the industry executives are directed in the Gartner post to make better algorithms to reduce bias; and to produce and act with restraint to reduce the public’s unease.
The post goes on to suggest executives “respond to jurisdictional difference by expanding the rights of people identified in images.” That is to say, they should ask who owns an image of a person, and the expression it depicts.
It is an intriguing question, one worth all the time spent on boilerplate queries and then some, and it is pushed around very briefly.
But it ignores the obvious. Which executives will implement a strategy answering that question to their satisfaction when, in the absence of standards or regulations, a competitor will scoop up sales by aiming lower?