Algorithmic Sabotage Research Group Asrg
Algorithmic Sabotage Research Group (ASRG) — Brief Report
Summary
Share your thoughts on our research and the implications for AI development. How can we work together to create more robust, secure AI systems? algorithmic sabotage research group asrg
Theorizing Algorithmic Sabotage: A collaborative document exploring prefigurative techno-political strategies. Algorithmic Sabotage Research Group (ASRG) — Brief Report
- Data Poisoning: This involves the intentional injection of "noise" into datasets. By feeding systems adversarial examples or conflicting data, the ASRG seeks to undermine the predictive accuracy of models. This is the digital equivalent of throwing a wrench in the gears, forcing the system to acknowledge its own fragility.
- Obfuscation and Chaff: Inspired by the use of "chaff" by aircraft to confuse radar, the ASRG develops tools that generate false digital footprints. If an algorithm tracks browsing history to build a profile, obfuscation tools flood the tracker with random, contradictory signals, rendering the profile useless. This reclaims privacy by denying the system the clarity it requires to function.
- Adversarial Fashion and Design: The ASRG researches physical-world sabotage, such as clothing and accessories designed to confuse facial recognition and computer vision systems. This "dazzle camouflage" for the digital age transforms the body into a glitch, making the wearer invisible or unrecognizable to the algorithmic eye.
The Algorithmic Sabotage Research Group serves as a vital provocateur in the tech landscape. They remind us that technology is a choice, not a natural law. Through their work, the "spanner in the works" becomes a tool for liberation, ensuring that as our world becomes more automated, it does not become less free. Data Poisoning: This involves the intentional injection of
: A computer science team at the University of Texas – Rio Grande Valley focusing on nanotechnology applications. Internet Research Task Force Manifesto on Algorithmic Sabotage Don't show me your AI. It is rude! - Tactical Tech
- Best practice: coordinated vulnerability disclosure to affected vendors, staged public disclosure after mitigations, and collaboration with defenders.
- Many legitimate research groups follow institutional review board (IRB) guidance and legal constraints; malicious use is a key concern.