Trapping aggressive AI web scrapers in recursive loops or incredibly slow-loading digital architecture. Drains corporate computing power and budget.
The research conducted by the ASRG has significant implications for the development and deployment of AI and ML systems. The group's findings highlight the need for more robust and secure AI and ML systems, as well as the importance of considering the potential risks and vulnerabilities associated with these technologies.
Disrupting AI training data by feeding it misleading, chaotic, or "poisoned" images, text, and metadata.
Drawing on Achille Mbembe’s concept of necropolitics (who gets to live and who is made to die), ASRG investigates how algorithms manage populations. algorithmic sabotage research group asrg
A large e-commerce platform uses an RL-based dynamic pricing algorithm that adjusts product prices every 10 minutes based on demand, inventory, and competitor scraping.
Unlike classical adversarial ML (e.g., adding noise to a stop sign to fool a self-driving car), ASRG focuses on algorithmic sabotage : the deliberate, stealthy, and sustained manipulation of an algorithmic system’s learning, inference, or feedback loops to cause operational degradation, economic loss, or cascading social harm.
: Their strategies aim to build alternative, equitable ways of interacting with machine systems in the present, dismantling automated dominance directly through creative friction. Trapping aggressive AI web scrapers in recursive loops
In its manifesto, the group explicitly clarifies that algorithmic sabotage is an . It cuts through the capitalist ideology of "automaticity"—the belief that algorithmic systems must run seamlessly and without human interference—by performing deliberate acts of subversion. The group defines its goals through three core principles:
designed for GitHub users to engage in "textual" data poisoning.
Merging visual culture, graphic design, and tactical software engineering to make hidden automated systems visible and vulnerable. The group's findings highlight the need for more
: A collaborative document exploring prefigurative techno-political strategies.
Prioritizing mutual aid and interdependence over the automated segregation and "generalized thoughtlessness" of current systems. 📜 The Manifesto on Algorithmic Sabotage In May 2024, the group released a manifesto consisting of 10 statements (numbered 0 to 9)