Autonomous weapons reinforce red lines for deterring hostilities.
Amatai Etzioni and Oren Etzioni, May-June 2017, “Pros and Cons of Autonomous Weapons”, Military Review, https://www.armyupress.army.mil/Journals/Military-Review/English-Edition-Archives/May-June2017/Pros-and-Cons-of-Autonomous-Weapons-Systems/
Why to begin. However, one should note that human-out-of-the-loop arms are very effective in reinforcing a red line. Declaration by representatives of one nation that if another nation engages in a certain kind of hostile behavior, swift and severe retaliation will follow, are open to misinterpretation by the other side, even if backed up with deployment of troops or other military assets. Leaders, drawing on considerable historical experience, may bet that they be able to cross the red line and be spared because of one reason or another. Hence, arms without a human in the loop make for much more credible red lines. (This is a form of the “precommitment strategy” discussed by Thomas Schelling in Arms and Influence, in which one party limits its own options by obligating itself to retaliate, thus making its deterrence more credible.) We suggest that nations might be willing to forgo this advantage of fully autonomous arms in order to gain the assurance that once hostilities ceased, they could avoid becoming entangled in new rounds of fighting because some bombers were still running loose and attacking the other side, or because some bombers might malfunction and attack civilian centers. Finally, if a ban on fully autonomous weapons were agreed upon and means of verification were developed, one could aspire to move toward limiting weapons with a high but not full measure of autonomy.