Zoom AI emotion recognition

Occurred: April 2022

Can you improve this page?
Share your insights with us

Plans by Zoom to develop an AI-based system that will monitor and detect the feelings and emotions of its users is drawing fire from US human and digital rights campaigners, ACLU, EPIC, Fight for the Future, and others.

According to an April 2022 report by Protocol, Zoom plans to introduce a suite of new features under the name Zoom IQ for Sales that use emotion AI to produce post-meeting transcripts and sentiment analysis to meeting hosts.

However, critics of emotion AI (or 'affective computing') argue the technology is often inaccurate as a person's facial expressions can be different to what they are feeling. 

It can also be discriminatory, with research showing the ways people express emotions vary significantly across cultures and situations. The technology is also said to be intrusive and easily misused. 

Plans by Intel to introduce emotion AI to schools are also proving controversial. 

Operator: Zoom Video Communications
Developer: Zoom Video Communications
Country: USA
Sector: Business/professional services
Purpose: Monitor & analyse emotion
Technology: Emotion recognition; Facial analysis; Voice recognition; NLP/text analysis
Issue: Accuracy/reliability; Bias/discrimination - race, ethnicity, disability; Dual/multi-use; Privacy; Surveillance
Transparency: Black box

Page info
Type: Incident
Published: April 2022
Last updated: May 2022