Open source intelligence and AI: a systematic review of the GELSI literature

Abstract

Today, open source intelligence (OSINT), i.e., information derived from publicly available sources, makes up between 80 and 90 percent of all intelligence activities carried out by Law Enforcement Agencies (LEAs) and intelligence services in the West. Developments in data mining, machine learning, visual forensics and, most importantly, the growing computing power available for commercial use, have enabled OSINT practitioners to speed up, and sometimes even automate, intelligence collection and analysis, obtaining more accurate results more quickly. As the infosphere expands to accommodate ever-increasing online presence, so does the pool of actionable OSINT. These developments raise important concerns in terms of governance, ethical, legal, and social implications (GELSI). New and crucial oversight concerns emerge alongside standard privacy concerns, as some of the more advanced data analysis tools require little to no supervision. This article offers a systematic review of the relevant literature. It analyzes 571 publications to assess the current state of the literature on the use of AI-powered OSINT (and the development of OSINT software) as it relates to the GELSI framework, highlighting potential gaps and suggesting new research directions.

Writing opacity: Going beyond pseudonyms with spirit portraiture

“…The description omits any information that could be used, with or without computation, to reidentify the subject. Granular details are substituted for encompassing terms like “vegan restaurant” or “childhood experiences.” In the former, although I loosely describe the setting of the restaurant, that this could be anywhere in four cities across Germany makes it almost impossible to identify, even using Open‐Source Intelligence (Ghioni et al., 2023). Where direct quotations are included, these are from exchanges that take place either before or after the recording.…”