On page 10 of ‘Who Is Responsible for Killer Robots? Autonomous Weapons, Group Agency, and the Military-Industrial Complex’, Taylor brings up List and Pettit’s argument that despite a certain group of individuals not having enough collective agency in a moral situation, one can still hold them responsible for their actions as if they did have agency. Subsequently, Taylor argues that one can only hold them accountable if doing so will push them toward developing agency. To this effect, Taylor presents an example of parents who may hold their children responsible, despite them not having the moral capacity to understand the repercussions of their actions. This can only be done if it encourages these children to develop a certain capacity for moral responsibility. For a military-industrial complex (MIC), the author says that this development of responsibility should ideally come about with organizational changes in how the MIC operates.

However, a problem I see here is that there seems to be an underlying premise that developing agency is a desirable attribute to avoid getting punished for not developing agency. People may not want to take the burden of moral responsibility on themselves, and therefore would not be incentivized to become moral agents. But despite not wanting to become moral agents, the cost of punishment drives them to do the same. For example, if a child does not like the feeling of being responsible for their toys, they will always act like a child when it comes to taking care of them (such as leaving them around). However, they will quickly learn that they are being punished for the same, and are forcefully driven to become moral agents despite not wanting to. I believe that the circumstances under which one becomes a moral agent would affect their behavior as a moral agent - you are not able to do a good job in a forced job. I believe this is an important aspect to be considered because the so-called quality and self-awareness of a moral agent would drastically affect the way they respond to the punishment and their ability to learn from their mistakes. Even in prisons, some people are more understanding of the crimes they’ve done than others and are therefore morally different people. Therefore, we cannot simply hold them accountable in the hopes of them developing agency because we cannot control the type of morality (so-to-say) that they would develop. In other words, we cannot guarantee that them developing morality will be a necessarily good thing because not all moral agents are the same.