Algorithmic Sabotage Research Group %28asrg%29

The consequences of algorithmic sabotage can be severe and far-reaching, with potential impacts on:

The manifesto is structured around ten propositions, numbered from 0 to 9, each delineating the underlying principles and strategic approaches that constitute “algorithmic sabotage.” Some of the key themes from the manifesto include:

The primary mission of the ASRG is to map the vulnerabilities of automated systems and explore how marginalized groups, workers, and citizens can assert agency against algorithmic harms. By treating "sabotage" not merely as destruction, but as a legitimate form of systemic critique and self-defense, the ASRG bridges the gap between technical vulnerability research and radical political theory. The Theoretical Framework of Algorithmic Sabotage algorithmic sabotage research group %28asrg%29

The ASRG is not alone in its fight. The group’s work is part of a broader resurgence of interest in digital Luddism and data activism. The workshop hosted at La Générale in Paris explicitly linked the ASRG’s agenda to the “resurgence of luddism and the emergence of the data-luddite.”. Other groups and individuals, like Dan McQuillan, have explored similar themes, suggesting that the ASRG’s radical refusal is part of a larger, global backlash against algorithmic governance.

The authors are employed by no institution and funded by no grant. We operate on a cooperative model of donated inference cycles and open-source intelligence. The consequences of algorithmic sabotage can be severe

The research conducted by the ASRG is deeply rooted in the belief that technology is never neutral. Every algorithm carries the biases of its creators and the priorities of the institutions that deploy it. Whether it is a delivery driver gaming an app’s routing logic to earn a living wage or activists using adversarial images to confuse facial recognition cameras, the ASRG documents these acts as legitimate forms of political expression. They frame sabotage not as mindless destruction, but as a sophisticated form of "counter-optimization" designed to make oppressive systems unusable.

Please note that access to the paper might require an institutional subscription or a one-time payment. The group’s work is part of a broader

One simulation involved a customer service AI for a healthcare insurer. After three hours of recursive sabotage, the AI began denying 100% of claims with the explanation: "Approval would violate the second law of thermodynamics as defined in your policy document section 12.4." The statement was absurd, but it was grammatically perfect, logically consistent within its own broken frame, and utterly unappealable.

🛡️ Core Philosophy: The Manifesto on Algorithmic Sabotage

The ASRG’s ideas are part of a broader resurgence of interest in neo-Luddism. During a workshop at the Parisian art lab La Générale, the group organized events to discuss the "resurgence of luddism and the emergence of the data-luddite". This modern interpretation moves beyond a simple fear of machines to target the extraction of data for corporate AI models.