AI in Warfare (Part 2): Peaceful Future or Slaughterhouse?
As mentioned in the first article, the war in Ukraine turned many eyes into the segment of military AI applications. Some that I listed are:
1. Development of Autonomous Weapons (LAWS) (autonomous tanks, swarming munitions/drones, etc.)
2. Military operations and their optimization (logistics, command and control, resources planning)
3. Platforms for intelligence collection and analysis (data from TikTok and Telegram to news reports and publicly available satellite imagery).
4. Detection of disinformation / or creation of it (posts and videos generated by troll farms on Social media — Twitter, Tik Tok, YouTube, Telegram, etc.)
Of course, many other segments could be added: planning and strategic decision making; improving tactical situational awareness; early warning systems; target acquiring; automatic point defense control; autonomous transport.
And this is implemented in the classical military operations, and it would be good to underline that AI-enabled cyber activities pose additional challenges and the battlefield. This alone could be a different topic to explore.
What is happening in practice? Ukraine‘s defense ministry began using Clearview AI’s facial recognition technology after the U.S. startup offered to uncover Russian assailants, combat misinformation, and identify the dead.
Clearview’s technology could be used to reunite refugees separated from their families, identify Russian operatives and help the government debunk false social media posts related to the war.
But not only Ukrainians are already using some AI-based solutions. Not surprisingly, Israel is also one of the frontrunners. “For the first time, artificial intelligence was a key component and power multiplier in fighting the enemy,” an IDF Intelligence Corps senior officer said.
The Israeli Defense Forces established an advanced AI technological platform that centralized all data on enemy groups in the Gaza Strip onto one system that enabled the analysis and extraction of the intelligence.
Collecting data using signal intelligence (SIGINT), visual intelligence (VISINT), human intelligence (HUMINT), geographical intelligence (GEOINT), and more, the IDF has mountains of raw data that must be combed through to find the key pieces necessary to carry out a strike.
They report that one strike, against a senior Hamas operative, was carried out with no civilian casualties despite being in a tunnel under a high-rise building surrounded by six schools and a medical clinic.
Hamas’s underground “Metro” tunnel network was also heavily damaged over the course of several nights of airstrikes. Military sources said they were able to map the network, consisting of hundreds of kilometers under residential areas, to a degree where they knew almost everything about them.
The mapping of Hamas’s underground network was done by a massive intelligence-gathering process that was helped by the technological developments and the use of Big Data to fuse all the intelligence. Once mapped, the IDF was able to have a full picture of the network both above and below ground with details, such as the depth of the tunnels, their thickness, and the nature of the routes. With that, the military was able to construct an attack plan that was used during the operation.
But all of this was still in the realm of Narrow and specialized AI systems. If we consider the possibility that somebody creates AGI, things change even more dramatically.
As some military experts predicted it would be simple for such an AGI to model the command and control networks, study the past decisions of the commanders, model their behavior, learn the SOPs, and be able to completely replicate and simulate the opposing C&C, unit movements, and strategies. Military training takes individual humans that are already pretty easy to predict, and brainwashes them into all thinking and acting the same, and acting according to the same SOPs. It makes them extremely predictable and vulnerable.
Such an AGI repurposed to Intel and Information Warfare (IIW) would be the most powerful weapon humans had ever developed. It could look days, weeks, or even months into the future and model millions of possible actions by the opposing forces, cull away the less promising and select the ones that make the opponent the weakest. Then it could simultaneously plot how to maneuver the opponent into the weakest positions by spoofing intel, feints in military maneuvers, and other IIW tactics, and how to marshal your units to be in a position to hit them the hardest when they were at their weakest.
Would we use those systems? Would we use other options that both AI and robotics could offer soon?
Would we want to tell a mother or father that their son or daughter is dead because it would have been unethical to send a robot into battle in their place? Many questions will have to be answered.