LAWs provide credible threat of retaliation, accelerating nuclear disarmament
Umbrello et. al 19 [Steven Umbrello, Phil Torres, and Angelo F. De Bellis. Umbrello works for the Institute for Ethics and Emerging Technologies and the University of Turin. Torres is an employee at the Project for Future Human Flpurishing. De Bellis is a Professor at the Universiy of Edinburgh. 01/24/2019, “The future of war: could lethal autonomous weapons make conflict more ethical?”, Journal of AI & Society, doi:10.1007/s00146-019-00879-x]
To begin, one of the most compelling reasons for opposing nuclear non-proliferation efforts is that the destructive potential of nuclear weapons increases the threshold of use (Jürgen 2008; Wilson 2012). Thus, only in extreme circumstances would rational actors deem their use to be either morally or strategically acceptable. This strongly contrasts with the case of LAWs, whose cost would be small compared to the cost of paying military personnel. Consequently, states could maintain stockpiles of LAWs that are far larger than any standing army. The low cost of LAWs would also make them more expendable than human soldiers (Jacoby and Chang 2008; Singer 2009a, b; Jenks 2010), and they could strike the enemy with greater precision than human scolders can currently achieve (Thurnher 2012; Ekelhof and Struyk 2014). These four properties—low cost, military effectiveness, expendability, and precision—could drive proliferation while lowering the threshold for use and, therefore, undermine geopolitical security. Incidentally, similar claims could be made about anticipated future nanotech weaponry (see Whitman 2011). The attractiveness of LAWs is apparent in the US’s use of “unmanned aerial vehicles” (UAVs, also known as “drones”) in Iraq and Syria. These semi-autonomous systems offer a cheap, effective, and relatively precise means for conducting surveillance and targeting enemy combatants [despite unsatisfied infrastructural needs to sustain the drone program] (McLean 2014). As a result, the US drone program has grown and the frequency of drone use against terrorist organizations like the (now-defunct) Islamic State has steadily increased in the past decade (Higgins 2017). Yet the proliferation of LAWs discussed in this paper is different in important respects from the proliferation of current UAV devices. LAWs are theoretically capable of becoming moral actors capable of making life and death decisions without human intervention. The absence of a human operator suggests that LAWs will be even cheaper than current UAVs and, as such, more vulnerable to proliferation. But this might not be undesirable given that, for example, ethical LAWs will—almost by definition—not serve to glorify or extend war efforts beyond the initial scope. Furthermore, UAVs still require human intervention and, as we will soon discuss, the emotional volatility of humans could lead to overspending and high death tolls. More generally speaking, the growing use of UAVs in conflict situations is consistent with a broader trend toward high-precision weaponry and away from larger, more destructive weapons like those in the world’s nuclear arsenals (Wilson 2013). There are some reasons for welcoming this shift. For example, the use of high-precision weapons like LAWs to achieve a state’s military objectives could reduce the probability and proportion of indiscriminate harm, thus violating the LoW and “rules of engagement” (RoE) less than might otherwise have been possible. Even more, the “ease-of-use” of LAWs that are fully autonomous could enhance the “balance of terror” that prevents conflict from breaking out by providing a credible means for retaliation: “If you strike me first, I will unleash a swarm of LAWs that devastate your infrastructure, poison your streams, set fire to your farms, destroy your armies, and assassinate your leaders.” The precision and effectiveness of LAWs could also accelerate the process of nuclear disarmament, seeing as the conception of LAWS regards them as agents capable of conventional weapons use rather non-conventional weapons platforms. First, consider that research on the potential climatic consequences of a nuclear war resulted in the replacement of MAD (“mutuallyassured destruction”) with SAD (“self-assured destruction”). The reason is that an exchange of nuclear weapons—even a regional one [citation]—could initiate a “nuclear winter” that causes global agricultural failures, widespread starvation, the spread of infectious disease, and other catastrophic sequelae that cannot be contained within national borders (Mills et al. 2014; Xia et al. 2015). Consequently, a nuclear war would all but guarantee the self-annihilation of states involved. As Seth Baum (2015) notes, though, LAWs could provide a kind of “winter-safe deterrence” by providing states with a credible threat of retaliation without the global catastrophic risks of nuclear conflict. Thus, LAWs could render the world’s nuclear arsenals irrelevant and, in doing so, lower the overall risk of human annihilation. Distinguishing valid targets As mentioned above and discussed further below, the most significant obstacle to the development and acceptance of LAWs is their current inability to reliably distinguish between valid and invalid targets. This is a crucial shortcoming because, as previously alluded, the Geneva Conventions require that attacks must be exclusively directed at military rather than civilian entities (United Nations 1979). As Article 51(3) of Sect. 1 of the Additional Protocol to the Geneva Convention states, “civilians shall enjoy the protection afforded by this Section, unless and for such time as they take a direct part in hostilities” (United Nations 1979). Complicating the situation further is an ambiguity about what exactly counts as a “civilian”—eg., some civilians engage in combative activities, some of which are not clearly overt, thus making it a matter of interpretation whether such individuals should fall within the “combatant” or “noncombatant” categories.