Algorithmic Sabotage Research Group %28asrg%29 -

: Sabotage is framed not as a hatred of technology, but as a form of counter-power and militancy absent from standard academic critiques.

The ASRG emerged from the broader field of . As governments and corporations increasingly rely on algorithms to manage welfare, policing, immigration, and finance, researchers have noted that these systems often perpetuate systemic bias and inequality.

Dr. Elara Venn had not slept in thirty-six hours. Not because she was overworked, but because she was afraid of what her dreams might calculate.

Techno-politics isn't about better code; it’s a political struggle. ASRG prioritizes radical feminist, anti-fascist, and decolonial perspectives to challenge "reductive optimizations".

Our next research phase (ASRG Cycle 5) will explore —using LLMs to produce appeal letters that are syntactically perfect but semantically absurd to the original classifier, forcing an endless loop of "escalate → deny → escalate." algorithmic sabotage research group %28asrg%29

The ASRG's core theoretical document is its Manifesto on “Algorithmic Sabotage” , a preliminary version first published in May 2024 that consists of ten propositions (numbered 0 through 9). It outlines the principles, strategies, and aesthetics of algorithmic sabotage. The manifesto begins with a stark epigraph from an anonymous partisan: “To create? No, to destroy, destroy and destroy again, whatever the strength left in these muscles allows. Because destruction is the power that is left. ... Only their destruction will last.”

The manifesto has been translated into multiple languages, including Greek and German, reflecting the group’s internationalist ambitions. An international call was also issued to translate the text into French, which took place during a workshop in Paris.

It remains unclear how much damage tarpits or other AI attacks can ultimately cause. Microsoft’s director of partner technology published a report detailing how leading AI companies were coping with data poisoning, one of the earliest AI defense tactics deployed. But as John Wiseman notes on his widely read blog, jwz , it is “very difficult to know whether that is effective because the only people who can answer that question are The Adversary.”

This article offers a comprehensive exploration of ASRG: its conceptual foundations, ten-point manifesto, expanding tactical toolkit, international workshops, and the broader context of data poisoning and adversarial AI research within which it operates. It also examines criticism and doubts about its effectiveness, while situating ASRG within the long historical lineage of technological sabotage. : Sabotage is framed not as a hatred

The ASRG categorizes sabotage into three distinct orders, ranging from individual resistance to systemic recalibration.

Declaring that the primary step in addressing technology is political rather than technological.

: Hosted on platforms like Our Collaborative Tools as a resource for prefigurative techno-political strategy.

The group published a manifesto containing ten statements (numbered 0 to 9) that outline the principles and aesthetics of their resistance. Artistic-Activist Resistance: Techno-politics isn't about better code; it’s a political

: Their work is deeply rooted in radical feminist, anti-fascist, and decolonial perspectives .

Rather than attempting to "fix" or optimize existing commercial machine learning models, ASRG explicitly rejects the foundational premises of mainstream AI development. The group’s philosophy shifts away from standard corporate "red teaming"—which it argues functions as free labor used to improve proprietary corporate tech—toward ideological, structural subversion. 1. Countering Algorithmic Authoritarianism

The is an artistic research collective and theoretical platform dedicated to investigating the politics of algorithms. Rooted in the traditions of tactical media, critical theory, and digital art, the group explores how "sabotage" can be used as a methodology to disrupt, expose, and challenge the power structures embedded within contemporary computational systems.

In short, the ASRG sees AI not as a technology to be tamed but as a weapon to be broken. While academic AI safety researchers focus on preventing rogue AI from harming humans, the ASRG focuses on enabling humans to harm AI—or at least to sabotage the systems that threaten their autonomy.