As a publication that relies on data collected in the Open Source, our writers have come across many people who mischaracterize the term Open Source Intelligence (OSINT). This article will attempt to explain the subtle nuances in the term OSINT and the major challenges analysts face when operating in the Open Source.
Open Source Intelligence (OSINT) is defined by the Joint Publication 2-0 as “Intelligence based on open source information that any member of the public can lawfully obtain by request, purchase, or observation.”
It is important to make the distinction between data, information, intelligence, and knowledge, all exploitable through Open Source.
Data is raw sensor inputs that are collected; this can include AIS data, university study findings, sentences written in newspapers, or listening to a speech. Information is data refined through background knowledge of a problem set. For example, data from an AIS collector will tell you a ship is declaring its location, name, speed, etc. This data becomes information when an analyst knows that that ship is Greek Flagged, leaving Iranian waters. That information becomes intelligence when it achieves relevancy to a decision maker with appropriate authority. For example, if an Office of Naval Intelligence analyst finds that this ship left from Bandar Abbas, carrying restricted Iranian fuel for an international buyer, that information does not become intelligence until it gets to the Commander of 5th Fleet in order to facilitate an action. Intelligence to one commander may only be background information to another, mission and relevancy is key. Finally, that piece of data does not become knowledge until staff and analysts combine that intelligence within the larger picture in order to accomplish a mission and provide a pattern of life which can perform follow on actions. OSINT is a significant tool in collecting data, but, by virtue of being publicly available, must be deeply refined before it is actually timely, relevant, and accurate intelligence.
Open Source Intelligence is described as “new” or “novel”. I would have to disagree. At the beginning of the Cold War, the Central Intelligence Agency absorbed the Foreign Broadcast Information Service (FBIS) in order to achieve infrastructure in developing data in the open source into actionable intelligence. This vignette demonstrates that OSINT has always been a tool for decision makers, but it was inherently categorized under HUMINT. Before the Digital Age and the Information Explosion of the early 2000’s, access to grey literature, newspapers, speeches, and all the other sources of OSINT required source placement in a foreign country. However, with the advent of the internet and globalization, intelligence agencies no longer needed to place sources in a nation’s capital or parliament to discern discussions or intent. That data and information became readily available online, thus the term OSINT was dubbed to describe what analysts were doing from afar. The best example that demonstrates the birth of OSINT from HUMINT due to globalization is North Korea where that evolution has not occurred. OSINT would not be considered a viable option for collections in North Korea because the trickle is so reduced. In order to collect Open Source data in North Korea, a human source placed there is still required due to the extreme control of the population’s interaction with the global community.
However, I would concede that Social Media Intelligence (SOCMINT) is truly a novel intelligence discipline that has emerged in the last 10 years. Social Media has evolved beyond its parent discipline, OSINT, in that social media is SOCIAL. Not only are people publishing vital data that can be used, but also providing their intent. Even more valuable is the collaborative nature of social media that allows discussions which analysts can snoop on from afar, now providing a web of data that can be exploited, much like being in a room full of people listening to their conversations. The war in Ukraine has solidified this fact as soldiers on both sides have gone to war with their smartphone in hand, providing a steady stream of geolocation data, combat footage, morale indicators, and much more. This data has provided decision makers with enough data to transfer precious ISR assets to collect in other areas, a strength of OSINT.
So, keeping all this in mind I categorize the two primary challenges to successful cultivation of Open Source Intelligence to be: the overwhelming volume of data and the vulnerability of that data to mis/dis information.
Intelligence analyst are increasingly being exposed to the use of artificial intelligence in order to counter the data explosion in the open source environment. 2.5 Quintillion Bytes of data are produced everyday, enough to completely inundate the intelligence community. So, by using tools such as Dataminr, Palantir, and ChatSurfer, analysts can set filters to sift through this data and pick out relevant points that provide indications and warnings. However, since this data is produced by anyone and everyone, it is reasonable to expect that adversary intelligence agents would produce disinformation to waste time and resources. OSINT can fill gaps that allow other expensive sensors to focus elsewhere. However, if a foreign actor publishes a data point that draws sensors to a ruse, we just lost time and resources that cost the enemy nothing.
For example, if a Syrian Defense Forces platoon records a video of an operation in one village, but saves the video and publishes it next week claiming to be in another location, multiple sensors would be allocated to locate those soldiers, who never had to get up from their chairs. This stands in contrast to the past where those soldiers would have had to create elaborate dummies or spend time and resources to deceive their enemies. That’s why intelligence analysts in the open source must corroborate data points with at least another source.
In summary: data collected in the Open Source is not OSINT, only careful cultivation of facts into relevant information to someone with authority to make decisions checks that box. Also, in order to effectively create OSINT, one must be cognizant of trustworthiness of sources.