Amazon Claims ‘Halo’ Device Will Monitor User’s Voice for ‘Emotional Well-Being’
September 1, 2020
Despite the exceptional privacy risks of biometric data collection and opaque, unproven algorithms, Amazon last week unveiled Halo, a wearable device that purports to measure "tone" and "emotional well-being" based on a user's voice. According to Amazon, the device "uses machine learning to analyze energy and positivity in a customer's voice so they can better understand how they may sound to others[.]" The device also monitors physical activity, assigns a sleep score, and can scan a user's body to estimate body fat percentage and weight. In recent years, Amazon has come under fire for its development of biased and inaccurate facial surveillance tools, its marketing of home surveillance camera Ring, and its controversial partnerships with law enforcement agencies. Last year, EPIC filed a Federal Trade Commission complaint against Hirevue, an AI hiring tool that claims to evaluate "cognitive ability," "psychological traits," and "emotional intelligence" based on videos of job candidates. EPIC has long advocated for algorithmic transparency and the adoption of the Universal Guidelines for AI.