Meta announced plans to install surveillance software on employee computers to monitor clicks, keystrokes, and all activity for the explicit purpose of training AI systems that will replace those employees. The system would capture comprehensive records of work behavior to generate training data for replacement AI.
The specific development is employers using workers to train their own replacement AI. This combines three elements: (1) comprehensive workplace surveillance (tracking all computer activity), (2) explicit replacement intent (the surveillance is specifically to replace workers), and (3) worker contribution without compensation (workers provide labor that trains the AI; the AI is company property).
The stability concern operates economically and socially. Economically, if companies can surveil employees to train AI that replaces them, labor becomes a non-renewable resource: workers are consumed to create replacement systems. Workers have no incentive to perform well (good performance accelerates their replacement) and no mechanism to negotiate (refusing to participate means being fired; participating means training your replacement). This creates perverse incentive: workers have incentive to deliberately underperform or obscure their processes so AI cannot learn from them.
Socially, it normalizes surveillance as condition of employment and explicit replacement as corporate strategy. Workers accept both surveillance (tracking every keystroke) and replacement (knowing they're training AI to displace them) as employment conditions. This represents dramatic shift in worker bargaining power: rather than automation being presented as gradual market development, replacement becomes explicit corporate policy announced to workers being surveilled.
The labor stability concern is immediate: if Meta does this, other tech companies will follow. Within years, employment in tech could require participating in your own replacement under comprehensive surveillance. Workers who refuse face unemployment; workers who accept face both surveillance and displacement. This creates labor market pressure that could affect other industries.
Historically, automation has displaced workers, but the displacement has typically been incremental and workers have been presented as needing to "upskill" and adapt. Meta's approach removes the pretense: replacement is deliberate policy, workers are source material for replacement systems, and the process happens under surveillance.
The data ethics concern is significant: workers' own problem-solving processes, their approaches to tasks, their work styles—all become company-owned training data. Employees are contributing intellectual property (how to do their jobs) to company-owned AI systems without compensation beyond salary.
Watch for: whether other tech companies implement similar surveillance and replacement training; whether employee organizations or unions respond; whether labor regulators challenge the practice; whether employees are given option to opt out; whether productivity or employee morale data is published showing impact; and whether tech worker recruitment or retention declines.