The Crowd-Sourced Intelligence Agency (CSIA) is an interactive artwork that allows users to perform the role of an intelligence analyst through an online interface. It is based on techniques known to be used by intelligence agencies from technical documents, Freedom Of Information Act (FOIA) requests, and leaked documents. The goal of the project is to help people understand both the effectiveness of Open Source Intelligence (OSINT) processing, and its impact on our privacy.
The CSIA has been shown internationally at art exhibitions, discussed at conferences, and has been used as the basis for academic articles and book chapters.
CSIA is taking some of the first steps towards creating tools for sous-dataveillance and counter-dataveillance. Interactive replications of intelligence techniques and technologies to allow participants to see some aspects of how dataveillance works in practice. By making abstract ideas about surveillance into concrete, CSIA fosters an informed public debate. The CSIA provides a visceral heuristic: as CSIA agents (users of the app) monitor their own posts and the posts of their friends, they can see how the automated processing changes, reinterprets, reframes, and recontextualizes their posts without needing a background in data science.
The CSIA features multiple supervised machine learning classifiers used for predictive policing. These classifying algorithms are modeled after the data collection processes described in technical reports and documents that have been leaked or released through Freedom of Information Act requests. All of the classifiers use the same Naïve Bayes algorithm, but each classifier is trained on a different corpus of textual data. By allowing users of the CSIA to compare the results from different classifiers, we hope to reveal the subjective nature of machine learning. Decisions about what data to include and omit from the training corpus directly determine the predictions the classifier will make, making predictive policing as much an art as it is a science.