About 15 years ago, the U.S. militarys elite counterinsurgency operators realized that the key to scaling up their operations was the ability to make sense of huge volumes of disparate data. From 2004 to 2009, Task Force 714 developed groundbreaking ways to sort and analyze information gathered on raids, which allowed them to exponentially increase the number of raids from about 20 a month to 300, SOCOM Commander Gen. Richard Clarke said Monday on a Hudson Institute broadcast.

The lessons from Task Force 714, which Clarke discusses further in this August essay, are now shaping how special operations forces uses AI in difficult settings, leading to such things as the Project Maven program.

Military leaders are fond of talking about how AI will accelerate things like predictive maintenance, soldier health, and increase the pace of battlefield operations. Were now seeing AI make some inroads into the command-and-control processes and its because of the same problem SOCOM hadan urgent need to make faster more effective decisions, said Bryan Clark, a senior fellow and director of Hudsons Center for Defense Concepts and Technology. Theyre finding in their wargaming that thats the only way U.S. forces can win.

And while service leaders have made a big show recently of how AI will accelerate operations in a high-end, World War III-style conflict, AI is more likely to see use sooner in far less intense situations, Clark said. AI might be particularly effective when the conflict falls short of war, when the combatants arent wearing identifiable uniforms. Such tools could use personal data the kind collected by websites and used by to sell ads targeted to ever-more specific consumer groups to tell commanders more about their human adversary and his or her intentions.

Take the South China Sea, where China deploys naval, coast guard, and maritime militia vessels that blend in with fishing boats. So you have to watch the pattern of life and get an understanding of what is their job on any particular day because if a fight were to break out, one, you might not have enough weapons to be able to engage all the potential targets so you need to know the right ones to hit at the right time, he said. But what if the conflict is less World War III than a murkier gray-zone altercation? What is the best way to defuse that with the lowest level of escalation? The best way to do that is to identify the players that are the most impactful in this confrontation and... disable them somehow, he said. I need enough information to support that decision and develop the tactics for it. So I need to know: whats the person on that boat? What port did they come out of? Where do they live? Where are their families? What is the nature of their operation day to day? Those are all pieces of information a commander can use to get that guy to stop doing whatever that guy is doing and do it in a way thats proportional as opposed to hitting them with a cruise missile.

You might think that grey zone warfare is a relic of the wars of the last ten years, not the modern, more technological competition between the United States, China, and Russia. But as the expansive footprint of Russia and China around the globe shows, confusing, low-intensity conflict, possibly through proxies or mercenary forces, should be an expected part of U.S., Chinese and Russian tension.

See the article here:

AI Is Reshaping the US Approach to Gray-Zone Ops - Defense One