Once confined to science fiction, autonomous and semi-autonomous systems are now active components of contemporary battlefields, all thanks to artificial intelligence (AI). From unmanned aerial vehicles (UAVs) to algorithm-driven decision-support systems, AI is accelerating the tempo of operations, improving the accuracy of gathering intelligence and enabling commanders to make faster and informed decisions under pressure.
Advocates hail AI as a force that is able to save lives and prevent strategic surprise, while critics warn of a future where autonomous weapons make decisions without human oversight.
UAVs (often known as drones) are the most visible markers of AI impact on warfare. UAVs have evolved into highly versatile platforms capable of surveillance, targeting, logistics support, and direct engagement in the late 20th century. Programming these UAVs are capable of providing real-time situational awareness, live feeds, engaging in enemy territories, and so on. This shift from tele-operation to increased autonomy reflects broader trends in AI deployment in combat systems.
Due to their quick reactions compared to humans, they offer potential tactical advantage. Yet, such autonomy raises concerns about the risk of unintentionally harming civilians.
Data acquired for training AI systems is also used to train recognition algorithms that can distinguish friendly from hostile forces, identify terrain features, and adapt to cluttered real-world conditions. These systems hinge on the quality of their training data, with incomplete or biased data leading to misclassification with severe consequences. Along with that, AI systems must cope with sensor noise and various such conditions where the cost of error is high.
It is not only in UAVs where AI is implemented on the battlefield. Intelligent systems now augment ground vehicles, naval assets, logistics chains, intelligence analysis, and even cyber defences, profoundly transforming the ways wars are waged.
Robotics and ground vehicles are also fed with AI data, which allows them to manoeuvre terrains, avoid obstacles, dispose bomb, and so on. Modern conflicts generate enormous volumes of data from satellite feeds and intercepted communications to social media posts and sensor logs.
In the seas, autonomous vessels are equipped with AI navigation systems that can patrol maritime zones with minimal human intervention. On the cyber front, defensive AI systems monitor network traffic for signs of intrusion. On the other hand, offensive cyber tools use machine learning to adaptively exploit vulnerabilities in adversary systems.
Despite their capabilities, the pervasiveness of AI across these domains introduces new challenges. Autonomous systems require robust cybersecurity to prevent hacking or manipulation.
Technology analyst Wes Roth highlighted in his X post how rapidly evolving AI capabilities could be repurposed in ways that undermine established norms of warfare and human oversight. His broader argument echoes ensuring strict ethical constraints on AI use in conflicts.
Others argue that AI, being a self-learning mechanism, is creating ‘killer robots’ which will corrode accountability and pose unacceptable risks to civilian lives. International humanitarian law emphasises by saying that current AI systems cannot fully replicate on the battlefields and that limited or biased datasets may misidentify targets, leading to severe consequences.
AI should not be misused, especially in harming someone else. But as we are currently in the AI Revolution, there is no stopping towards the various fields which can be benefited by using AI. If used on the battlefields, it should be noted that the datasets fed into the AI of war machines and vehicles should be kept updated all the time to prevent the loss of innocent human lives.
