So it learns what content tends to get algorithmically boosted, hunts down content that could fit those traits, then let's the DoD know to "deploy countermeasures". They want total narrative control; along with their new classifications of what constitutes terrorist activity, like criticising environmentalism or government spending, they'll be targeting anything that isn't mundane pop culture drivel. I guess this just means they are getting an AI boost to censorship they were already doing. Crazy how open they are about the DoD taking an active role in censorship.