Military and Public Attitudes Toward Autonomous Weapons: Stage-Specific Drivers and Value-Based Moderation

Warfare has always been shaped by the tools of the day, from the first stone projectile to the complex machinery of the twentieth century. Today, a new kind of tool is emerging: weapons that can choose their own targets and decide when to strike, guided by artificial intelligence rather than a human finger on a trigger. These systems, often called autonomous weapons, raise a profound question that goes beyond engineering: at what point does a machine become too independent to be trusted with life and death? The debate is not just about whether these machines work, but whether people are willing to let them work. It hinges on two distinct ideas. First, there is the idea of building the technology, which feels like a strategic necessity to many. Second, there is the idea of actually using it in a real fight, where the consequences are immediate and irreversible. Understanding the difference between supporting a new invention and supporting its use in battle is crucial, because the people who might have to live with the results—the soldiers on the ground and the citizens who fund the wars—often see these two stages very differently.

A team of researchers in South Korea set out to map these differences by asking two very different groups of people about their views on autonomous weapons. They surveyed 1,222 members of the general public and 155 active-duty military personnel. The goal was to see if the people who might actually have to use these weapons felt the same way as the rest of the population, and to understand what factors, from personal values to technical knowledge, drive those feelings. The researchers did not just ask for a simple "yes" or "no." Instead, they broke the question down, asking separately about the development of the technology and its deployment in combat. They also looked at how much people knew about artificial intelligence, what they believed about the ethics of war, and what they valued most in society, such as economic stability versus personal freedom.

The results revealed a striking split in perspective. The military personnel were more enthusiastic about the idea of developing autonomous weapons than the general public was. This makes sense, as soldiers often view new technology as a way to gain a tactical edge or protect their own lives. However, when the question shifted to actually using these weapons in a real battle, the military group became significantly more cautious than the public. While the public was relatively more open to the idea of machines making lethal decisions, the soldiers, who understand the fog of war and the weight of responsibility, were far less willing to hand over that power. It is as if the military sees the weapon as a useful tool to build, but draws a hard line at letting that tool pull the trigger without a human standing right there.

This caution among the soldiers was not random; it was deeply tied to their sense of ethical responsibility. The study found that the strongest factor influencing whether anyone would accept these weapons in combat was their belief that the machines could follow the ancient rules of war. These rules require that a weapon can tell the difference between a soldier and a civilian, that the force used is proportional to the threat, and that every effort is made to avoid unnecessary harm. When people felt confident that an autonomous system could meet these strict ethical standards, they were more likely to support its use. But when they doubted the machine's ability to make those moral distinctions, support dropped sharply. For the military, this doubt was even more pronounced, suggesting that those closest to the reality of combat are acutely aware of the difficulty of programming a machine to make split-second moral judgments.

Perhaps the most surprising discovery concerned what happens when people know more about the technology. A common assumption is that if people understand how artificial intelligence works, they will be more comfortable with it. The researchers found the opposite to be true regarding the use of these weapons. As people's knowledge of artificial intelligence increased, their support for using autonomous weapons in battle actually decreased. This suggests that a deeper understanding of the technology does not lead to blind optimism. Instead, it leads to a more informed skepticism. Those who know how these systems work, how they learn from data, and where they might fail, are more likely to see the risks and uncertainties involved in letting them make life-or-death decisions. Knowledge, in this context, acts as a brake rather than an accelerator.

The study also showed that a person's core values play a major role in how they react to this new knowledge. People who prioritize material concerns, such as economic stability and national security, tended to maintain their support for these weapons even as they learned more about them. Their focus on safety and strength seemed to outweigh their ethical worries. In contrast, people who prioritize post-materialist values, such as human rights, freedom of expression, and environmental protection, reacted very differently. As their knowledge of artificial intelligence grew, their support for using autonomous weapons plummeted. For this group, the ethical implications of handing over lethal power to a machine became more salient the more they understood the technology. Their values made them more sensitive to the potential for error and the moral cost of automation.

The researchers also looked at the specific jobs these weapons might do. Support for autonomous weapons was not a blanket approval; it depended heavily on the mission. Both the public and the military were more willing to accept these systems for tasks like surveillance, reconnaissance, and gathering information. However, support dropped significantly when the mission involved direct attacks, striking high-value targets, or engaging in aggressive combat. This indicates that people do not reject the technology entirely, but they are highly selective about where and how it is used. They are comfortable with machines watching and listening, but far less comfortable with machines striking.

Ultimately, the study paints a picture of a society that is not simply for or against autonomous weapons, but is navigating a complex landscape of trust, risk, and ethics. The path to deploying these systems is not just a matter of making them work better; it is a matter of convincing people that they can be used responsibly. The gap between the military's desire to develop the technology and their hesitation to use it suggests that the biggest hurdle is not technical capability, but ethical legitimacy. If the people who have to live with the consequences of war cannot be convinced that these machines will respect the rules of human conflict, then the technology may remain in the laboratory, no matter how advanced it becomes. The findings suggest that the future of these weapons depends less on the code written by engineers and more on the moral boundaries drawn by the people who must live with the results.