"The OSINT center of gravity resides in human operators, and not in any machine or tool. Machine based collection and processing will augment highly skilled human operators orchestrating the compilation of multiple streams of open source information to create high-value intelligence reports, cross-cue other collection platforms, and feed the all source analysis process."
— Department of Defense OSINT Strategy 2024-28

Open-source intelligence, or OSINT, is intelligence derived from information collected through publicly available sources and then analyzed and shared to meet specific intelligence needs. Worldwide, more than 400 million terabytes of data are generated every day, providing massive amounts of potential open source intelligence.

Perhaps this is why the DoD, in its OSINT Strategy 2024-28, has said it aims to “elevate OSINT as a core intelligence discipline that provides high-value unclassified reporting, while simultaneously serving as a catalyst for enriched all source analysis and a valuable cueing mechanism for the other intelligence arts.”

From social media, to news, photos and videos, there is an ever-growing mound of data available for use in intelligence analysis. The pace of data creation is going to continue to expand faster than our ability to mine it. Tools – especially artificial intelligence and machine learning solutions – are accelerating our ability to find proverbial needles in haystacks, but can be limited when it comes to the innate human understanding that is often required to turn information to intelligence.

AI is incredibly powerful; it is able to create efficiency by ingesting and prioritizing data for analysis. AI is great for parsing through large amounts of information posted on the internet. Currently, the optimal way to use AI for analysis would be for tasks that condense information and produce a product for human analysis and evaluation of information that is not black and white.

Humans Understand Nuance

Machines can do a lot, are getting smarter all the time and have an increasingly important role in the intelligence analysis space, but it is important to keep a human in the loop. There are millions of people on the planet who speak the same language, but dialects differ from country to country and from region to region. A word that’s positive in one spot on the globe can mean something completely different in another.

Sarcasm is also famously hard for machines to understand, especially in a machine translation context. Think about domains like finance or banking or sports – in these domains, a “surge” is a good thing but in a medical context, a surge can mean a spike in the number of cases of a disease or an influx of emergency room patients. A machine can translate but a human can interpret – there’s a big difference.

Analysts Maintain a Healthy Sense of Skepticism

Intelligence analysts must maintain a healthy sense of skepticism. This is especially important in today’s world where there’s an immense amount of disinformation or misinformation intentionally created and spread. Machines translating or assessing content in this environment will not be naturally skeptical. They’ll perform in their binary, black and white nature, leaving little room for gray.

Human-in-loop intelligence is important in every intelligence domain, but especially in OSINT given the sheer variety of sources, languages and content-creators – some of them with nefarious motives – sending information into the ether.

Asking Questions Borne of Curiosity

Human OSINT analysts must, and often do, ask questions. They have a natural curiosity and a desire to build on their experiences. They are methodical, measured and they're able to question their own biases and spot disinformation or misinformation.

The more data we continue to generate the more gaps we will have in our ability to assess it all. This is precisely why we need the essential combination of intelligence tools and people with the skills necessary to analyze it all.

GDIT and the International Spy Museum have partnered to produce the museum's first-ever digital exhibit, Open Source: Ukraine & the Intelligence Revolution. This exhibit is a real-time look into the practice of extracting meaningful intelligence from the sea of data published every second.