Autonomous drone swarms

The ongoing conflict in Ukraine has served as a testing ground for the latest drone technology, with experts expressing concerns over the increasing role of artificial intelligence (AI) in decision-making for warfare. In terms of scale and variety, the use of drones in Ukraine is unprecedented, with tanks, trench warfare, and artillery still dominating the conflict. However, there are few international legal norms in place to define the extent to which human involvement is required for the use of drones. Some manufacturers, such as Russia's Lancet and the US-provided Switchblade, claim their machines now have autonomous or semi-autonomous capabilities.

As AI for drone technology continues to develop, the concern is that it will be used in the field before international norms are in place. The rush to defend Ukraine has accelerated the discussion around weaponizing AI, as evidenced by Ukraine's fundraising for what it calls "the world's first naval fleet of drones" and Russia's rush to deploy systems in the field.

One area of concern is the use of "swarm" technology, in which multiple drones work together to attack a target. Currently, these drones are guided at launch by a human operator, but true swarming capabilities are on the horizon. In January 2022, Raytheon technology was used in a DARPA exercise in which just one human operator controlled 130 adapted commercial off-the-shelf drones as they swarmed autonomously and surveilled an area.

As AI technology continues to evolve, the concern is that decision-making for drone strikes will become increasingly autonomous, with human operators becoming more reliant on the machine's prompts. This presents unique problems, as human nature tends to trust outputs presented by computer-assisted systems more than our own judgment.

The military mission steps for facilitating an automated cyberkillchain would involve the integration of artificial intelligence (AI) and unmanned aerial vehicles (UAVs) into the traditional decision-making process. The first step would involve the use of AI-enabled surveillance and reconnaissance capabilities to identify and track potential targets. This data would then be fed into an AI-enabled decision-making system, which would analyze the information and determine the best course of action.

Once a target has been identified, the UAVs would be deployed to carry out the attack. These UAVs would be equipped with autonomous or semi-autonomous capabilities, allowing them to carry out the attack without human intervention. This could include the ability to evade detection, navigate complex terrain, and strike targets with precision.

The final step in this process would involve the use of AI-enabled post-strike analysis to evaluate the effectiveness of the attack and identify any potential improvements for future missions.

In this scenario, conventional decision making would be integrated with the attack steps to autonomous killing at scale. The human operators will provide the initial guidance and direction, but the AI system will take over the decision-making process once the target is identified. The human will be in the loop of control, meaning the drone can fly autonomously, circling an area until a target appears, but are otherwise operated by a human.

As the technology improves, the human role in decision-making may be reduced, with the AI system taking on more autonomy. This is where the concern of "on" the loop of control, meaning the human will not be involved in the decision-making process, and the AI system will have the final say on whether a target should be attacked or not.

The future of warfare may see the emergence of what experts call the "cyberkillchain," a process in which AI-controlled drones operate independently, identifying and attacking targets without human intervention. As AI technology continues to evolve, the concern is that the decision-making process for drone strikes will become increasingly autonomous, with human operators becoming more reliant on the machine's prompts, potentially leading to a future in which AI makes the decision of whether or not humans live or die.