Activists are warning that a new feature from Zoom Video Communications analysing emotion and engagement could introduce bias and risk violating personal privacy.
The maker of the ubiquitous video conference technology last month announced Zoom IQ — a post-call analysis of speaker sentiment powered by artificial intelligence. Designed for salespeople, the tool provides metrics like engagement or patience in an effort to improve future communication and deal-making.
More than 25 organisations, including the American Civil Liberties Union and the Electronic Privacy Information Center, asked the company on Wednesday to abandon the feature. In a letter to CEO Eric Yuan, the group said that emotion analysis isn’t backed up by science and presents the potential for discrimination and the release of personal data.
“There is zero reliable evidence that a machine can accurately assess someone’s emotional state and a lot of evidence that one-size-fits-all assumptions about ‘normality’ don’t mirror human diversity,” Tracy Rosenberg, of the citizens’ coalition Oakland Privacy, said in a statement.
The group is also launching a site where members of the public can virtually sign a petition urging the company to drop the new feature. Fight for the Future, which is organising the campaign, has also coordinated online protests against corporate use of facial recognition, and in favour of tech regulations like US antitrust bills and net neutrality in 2012. A commitment from Zoom is requested by 20 May.
Zoom has responded in the past to public pressure on product decisions. Early in the pandemic, the company axed a feature that tracked meeting attendee engagement, such as when they clicked outside the chat window. Later, it suggested that end-to-end encryption would not be available to free users, before reversing course. — Brody Ford, (c) 2022 Bloomberg LP
