The U.S. has warned for weeks about Russia’s possible invasion of Ukraine. It is threatening retaliation if it does. Eight years after Russia’s invasion of eastern Ukraine and the invasion of Crimea, Russian forces have begun to mobilize along Ukraine’s borders.

The U.S. and NATO member countries monitor Russia’s activities to determine the best policy responses. They no longer rely solely on multimillion-dollar spy satellites or ground spies for timely intelligence.

Big data, social media, smartphones and low-cost satellites are prominent in today’s intelligence analyst toolkit. Twitter scraping has been deemed as crucial as any other intelligence analysis tool. These technologies allow news organizations and armchair detectives to track the action and provide analysis.

Governments still carry out sensitive intelligence-gathering operations with the help of extensive resources like the U.S. intelligence budget. However, there are vast amounts of valuable information that is publicly accessible. Not all of it is available to governments. Private companies can now operate drones and satellites at a fraction of the cost. Nearly everyone has a smartphone that supports advanced video and photo capabilities.

As an intelligence scholar and information operations scholar, I study the technology that produces massive amounts of intelligence data and help to find relevant information.

Open-source intelligence

The truth about Russia’s military posture can be accessed via the internet through information gathered by individuals and commercial companies. Commercial imaging companies are posting images of Russia’s military forces. Numerous news agencies monitor and report on the situation. TikTok users are posting videos showing Russian military equipment on rail cars. These rail cars are allegedly being used to augment the forces in Ukraine. Internet sleuths track this information flow.

Intelligence professionals will find this democratization of intelligence collection a boon. Instead of relying on expensive sensors or classified systems high up or above the earth, government analysts can now make intelligence assessments using information from the web.

It cannot be easy to sort through the terabytes of publicly available data to find relevant information. The task is more difficult because some data may have been intentionally altered to deceive.

Open-source intelligence is a practice. The U.S. director for national intelligence describes Open Source Intelligence as the collection, evaluation and analysis of publicly accessible information. Information sources include news reports, social media posts, YouTube videos, and satellite imagery provided by commercial satellite operators.

OSINT agencies and communities can use many free tools. Analysts can use the tools to create network charts, such as criminal organizations, by searching publicly available financial records.

Private investigators use OSINT to support law enforcement, corporate, and government needs. Lazy sleuths have used OSINT to expose corruption and crime to the authorities. OSINT can meet the majority of intelligence requirements.

Machine learning for intelligence

OSINT is not a perfect tool or practice. It contributes to the overload of information intelligence analysts must deal with. Intelligence analysts are often in reactive mode, trying to understand a continuous stream of confusing raw data and information.

Machine Learning is a set of methods that allow computers to recognize patterns in large quantities of data. This is especially useful for processing OSINT information, such as photos and videos. The speed of computers in sifting through large data sets makes it easier to use machine learning tools and techniques that will optimize the OSINT process.

Computers can identify patterns to help them evaluate the information for credibility and deception and predict future trends. Machine learning, for example, can be used to determine whether data was created by humans, bots, or other computer programs and whether it is genuine or fraudulent.

Machine learning is not a perfect tool, but it can help to evaluate the probability of certain outcomes. While no one can read the mind of Vladimir Putin, the OSINT/machine learning combination could be used to help analysts evaluate how, for instance, a Russian invasion in Ukraine might unfold.

Technology has created a tsunami of intelligence data. However, technology is making it easier to extract meaningful data from that data to aid human intelligence analysts in putting together the big picture.