In a decision which seems to be inspired by the racially-charged protests going on around the world, IBM has announced that it is suspending work on general purpose facial recognition or analysis software.
In a letter sent to congress on Monday, IBM CEO Arvind Krishna said the company will no longer pursue facial recognition technology.
“IBM no longer offers general purpose IBM facial recognition or analysis software. IBM firmly opposes and will not condone uses of any [facial recognition] technology, including facial recognition technology offered by other vendors, for mass surveillance, racial profiling, violations of basic human rights and freedoms, or any purpose which is not consistent with our values and principles of Trust and Transparency.
“We believe now is the time to begin a national dialogue on whether and how facial recognition technology should be employed by domestic law enforcement agencies,” he said.
Facial recognition software has been surrounded by controversy from the first time it was trialed. In 2018, research conducted by Joy Buolamwini and Timnit Gebru found that most of the systems were biased. The finding spurred further criticism of the algorithms and efforts to fix the shortfalls of the system that resulted in bias.
The Verge analysis of the entire system found more loopholes in other areas including privacy. In December 2019, the National institute of Standards and Technology discovered “empirical evidence for the existence of a wide range of accuracy across demographic differences in the majority of the current face recognition algorithms that were evaluated,” part of it, violation of privacy that has become an everyday story of technology.
Most of the companies selling the facial recognition software have been under scrutiny. Even though not all the companies were involved in the studies conducted, many that were involved had cases to answer. For instance, Clearview AI came under scrutiny when it was discovered that its facial recognition tool, built with more than 3 billion images compiled partly from social media sites is being used by private organizations and law enforcement agencies.
IBM attempted to help with the issue of bias in facial recognition through the public data set it released in 2018. The data showed that commercial systems are more accurate if you’re a white male. This is because of lack of system technique to accommodate diversity; people of color appear less frequently than others, limiting their chances of recognition.
While IBM was still making an effort to clear the issue of bias, it was caught in a privacy controversy. The company was found sharing a separate training data set of nearly one million photos taken from Flickr without the consent of the subjects in 2019.
Other companies were also getting into trouble for the use of facial recognition technology. Facebook faced a class-action lawsuit for unlawful use of facial recognition technology that it settled with $550 million in January.
In 2018, the American Civil Liberties Union found that Rekognition, facial recognition software sold by Amazon, incorrectly matched 28 members of Congress to faces collected from 25,000 public mugshots.
IBM also considered the issues of racial profiling by law enforcement agents. As part of the reason it’s discontinuing its work on facial recognition technology, it urged the government to reform the police and address the ever increasing cases of racial injustice.
In the letter, Krishna quoted the letter Thomas J. Watson Jr., the 1953 president of IBM, wrote to all employees:
“…Each of the citizens of this country has an equal right to live and work in America. It is the policy of this organization to hire people who have the personality, talent and background necessary to fill a given job, regardless of race, color or creed.”
He added: “congress should bring more police misconduct cases under federal court purview and should make modifications to the qualified immunity doctrine that prevents individuals from seeking damages when police violate their constitutional rights.
“Congress should also establish a federal registry of police misconduct and adopt measures to encourage or compel states and localities to review and update use-of-force policies.”
With abundant evidence of inconsistencies of facial recognition technology, especially the lapse in recognition of people of color, congress is likely going to act on IBM’s letter. Other companies that may still have the intention to continue their work with the technology will eventually be stopped or face tighter rules.