In the burgeoning field of Machine Learning (ML) security, most research focuses on defense: robust aggregation, differential privacy, adversarial training, and anomaly detection. A smaller, more provocative, and increasingly vital niche focuses on offense—not to break systems for malice, but to understand their catastrophic failure modes. At the radical fringe of this offensive security research lies the hypothetical (and increasingly real) collective known as the Algorithmic Sabotage Research Group (ASRG).
As AI models become increasingly inscrutable, the ASRG's work serves as a "collective counter-intelligence". They advocate for: Communal Constraints:
Theorizing Algorithmic Sabotage: A collaborative document exploring prefigurative techno-political strategies. algorithmic sabotage research group asrg
For further investigation into these perspectives, public documentation and collaborative platforms hosting these discussions can be found through digital research archives and academic databases focused on media theory and tactical media history. Drop #17. Manifesto On Algorithmic Sabotage
: Their work is deeply influenced by radical feminist, anti-fascist, and decolonial perspectives, which challenge the "reductive optimizations" of modern algorithms. Resistance as Creativity At the radical fringe of this offensive security
As AI continues to permeate various sectors, the work of ASRG and similar research groups becomes increasingly critical. Future directions for ASRG include:
The Mission of ASRG
Building networks of solidarity that algorithms—by their very design—cannot compute or categorize.