Meta exploits workplace surveillance to extract unconsented labor for AI training, deepening digital feudalism and precarious labor

Indigenous epistemologies reject the commodification of labor and data, framing knowledge as relational and communal rather than extractable property. The Māori principle of 'kaitiakitanga' (guardianship) asserts collective stewardship over data, directly opposing Meta’s extractive model. Similarly, Andean 'ayni' (reciprocity) principles critique uncompensated labor extraction as a violation of social contracts. These frameworks expose the ethical bankruptcy of treating human cognition as raw material for AI.

The extraction of workers’ cognitive labor mirrors historical patterns of enclosure, from the Enclosure Acts to the commodification of factory labor during the Industrial Revolution. The 'scientific management' (Taylorism) of the early 20th century sought to optimize worker efficiency without compensation, foreshadowing today’s AI training pipelines. The rise of 'digital Taylorism' in platform economies replicates these dynamics, where algorithms dictate labor while corporations capture all surplus value.

In China, platform workers have organized under the banner of 'digital labor rights,' drawing on Maoist-era labor traditions to demand fair compensation. European labor movements, influenced by Nordic models, have pushed for 'data dividends' to compensate workers for data extraction. Latin American cooperatives, like Brazil’s 'Catarse,' demonstrate alternative models where workers control their data and AI outputs collectively.

Neuroscience confirms that cognitive labor (e.g., keystrokes, attention patterns) is metabolically costly, yet Meta treats it as zero-cost input for AI training. Labor economics research shows that uncompensated digital labor depresses wages and increases precarity, particularly in Global South outsourcing hubs. Studies on algorithmic management reveal how surveillance reduces worker well-being and productivity, contradicting Meta’s efficiency claims.

Artists like Hito Steyerl critique the 'poor image'—a metaphor for how digital labor is reduced to raw material, devoid of human dignity. Indigenous artists such as Jaune Quick-to-See Smith depict data extraction as a form of cultural violence, echoing the dispossession of land and knowledge. Spiritual traditions like Buddhism warn against 'attachment to outcomes,' a principle violated when corporations monetize human attention without consent.

If unchecked, this model could lead to a 'digital feudalism' where corporations own the means of cognitive production, and workers are perpetually indebted to platforms. Scenario modeling suggests that without regulation, AI training costs could collapse, accelerating monopolization and reducing innovation diversity. Conversely, a 'data commons' model could democratize AI by compensating workers and redistributing surplus value.

Black and brown workers in Global South call centers bear the brunt of uncompensated digital labor, yet their voices are excluded from tech policy debates. Women, who dominate platform-based care work (e.g., content moderation), face heightened surveillance and exploitation under Meta’s model. Disabled workers, who rely on digital platforms for accessibility, are disproportionately impacted by opaque data extraction practices.