The Quiet Resistance: Understanding Algorithmic Sabotage at Work
In the United States, the Computer Fraud and Abuse Act (CFAA) criminalizes "knowingly caus[ing] the transmission of a program, information, code or command, and, as a result of such conduct, intentionally caus[ing] damage, without authorization, to a protected computer". In theory, data poisoning and systematic algorithm manipulation could be prosecuted under this statute. In practice, enforcement remains rare and legally uncertain. A district court has ruled that creating fake accounts to study algorithmic bias does not violate the CFAA, but this does not shield individuals from liability under other laws.
Protect the core recommendation/classification algorithm from manipulation by detecting and quarantining "sabotage" inputs (adversarial examples or poisoned data).
# Reshape for single sample prediction if input_data.ndim == 1: input_data = input_data.reshape(1, -1) algorithmic sabotage work
Organizations cannot solve algorithmic sabotage through harsher surveillance; workers will always find a new loophole. Instead, the solution requires a fundamental shift in how workplace technology is designed and implemented. Human-in-the-Loop Management
Ultimately, algorithmic sabotage highlights a growing friction between human intuition and mathematical efficiency. As long as management relies on opaque code to control labor, workers will continue to find the "ghost in the machine"—turning the algorithm’s own logic against it to protect their livelihood.
Algorithms often set optimization goals based on mathematical ideals rather than human physical limitations. Workers manipulate data to lower these impossible benchmarks. A district court has ruled that creating fake
Algorithmic sabotage is rarely just about laziness; it is often a rational response to surveillance and disempowerment. 1. Protection Against Unfair Evaluation
The tactics of algorithmic sabotage are as diverse as the industries they target, ranging from subtle forms of non-compliance to sophisticated attacks on core data infrastructure.
Author’s Note: The tactics described in this article are based on ethnographic research, leaked internal documents, and anonymous interviews with gig workers. The author does not endorse time theft but recognizes it as a sociological inevitability under algorithmic management. Instead, the solution requires a fundamental shift in
Algorithms frequently reproduce historical biases and lack the contextual nuance of a human manager. Employees may engage in resistance to protect themselves and colleagues from unfair AI-driven performance evaluations. 2. Loss of Autonomy and Dignity
Simple scripts running in the background that simulate typing, ensuring that automated activity trackers register continuous productivity. 3. Gamifying the System
Modern management relies heavily on software to track, evaluate, and direct human labor. From algorithmic scheduling in retail to automated keystroke logging in remote tech jobs, artificial intelligence has become the new middle manager. However, workers are not passive inputs in an equation. As automated systems squeeze labor for maximum efficiency, a new form of resistance has emerged: .