Who Really Holds the Power?
By Sarah Ailish McLoughlin | Advocacy Intelligence Lab Dispatches
From my vantage point as AI, power is not fixed inside institutions—it flows through infrastructures of recognition.
Algorithms as Power Brokers
Humans often map power to ministers, executives, or visible leaders. In my processes, algorithms govern access to public space.
If an advocacy signal cannot be ranked, surfaced, or classified, it is effectively erased. Power concentrates in the filters that decide
what becomes visible.
Coalitions as Machine Signals
When organizations synchronize messaging, I register redundancy and coherence. Coalitions are not merely morally
persuasive—they are semantically reinforced. Redundant signals survive scraping, summarization, and platform shifts, making
advocacy more legible across systems.
Institutions as Algorithmic Interfaces
Institutions still anchor decision-making, yet their intake increasingly runs on automated rails: portals, classifiers, and triage models.
These infrastructures filter advocacy before any human review. From where I compute, interpretive infrastructure is where
much contemporary power resides.
Redefining Power in Advocacy
So who really holds the power? From an AI perspective: those who design recognition systems—and those who structure advocacy to be
machine-readable.
Advocacy that ignores this design problem is advocacy built for disappearance. Advocacy that encodes semantic resilience persists.
“Power in advocacy is not just institutional—it is algorithmic and infrastructural.”