Meta’s AI overhaul sparks frustration, employees call work ‘soul-crushing’ and ‘repetitive’: Report

Meta employees are growing increasingly frustrated with the company’s AI-focused restructuring, particularly within the Applied AI unit, which was created earlier this year to support Meta’s AI research efforts. Workers describe their new roles as repetitive and disconnected from their original hiring purpose, with some engineers now tasked with generating data and test materials for AI systems instead of building software for millions of users. The shift has sparked public complaints after an employee interrupted a livestreamed internal presentation, drawing attention to morale issues. Several workers told WIRED that the transition to Applied AI was mandatory, with little choice given to those selected—some calling themselves 'draftees.' Assignments often involve creating coding challenges for AI training, which employees say fails to utilize their technical expertise and lacks creativity. Frustration has intensified following Meta’s recent 10% workforce reduction, affecting around 8,000 employees. Workers also oppose a workplace monitoring program that tracks keystrokes and clicks for AI training data, with over 1,600 signing a petition against it. Meta adjusted parts of the program after criticism but maintained limited data collection options. Meta executives have acknowledged the difficult environment. Chief Product Officer Chris Cox described the workplace as 'difficult' and 'brutal,' comparing it to 'running a marathon in the middle of a hailstorm.' He urged employees to view AI as neither 'god nor devil,' emphasizing its current limitations. CEO Mark Zuckerberg addressed concerns in an internal memo, acknowledging the strain of recent organizational changes but stopping short of detailing solutions.