human-out-of-the-loop

Human-out-of-the-loop describes a level of autonomy in weapons systems where, once activated, the machine selects and engages targets without further human intervention. It is typically contrasted with two other modes:

• Human-in-the-loop: the system can select targets but cannot fire without explicit human authorization.
• Human-on-the-loop: the system can select and engage autonomously, but a human supervisor monitors operations and can override.

The taxonomy is widely used in policy debates over lethal autonomous weapons systems (LAWS), particularly within the UN Convention on Certain Conventional Weapons (CCW) Group of Governmental Experts, which has met in Geneva since 2017 to consider regulation. The phrase also appears in the US Department of Defense's directive on autonomy in weapon systems (DoD Directive 3000.09, originally issued 2012 and updated in 2023), which requires that autonomous and semi-autonomous weapons allow commanders and operators to exercise "appropriate levels of human judgment over the use of force."

Critics — including the Campaign to Stop Killer Robots, the International Committee of the Red Cross (ICRC), and UN Secretary-General António Guterres, who in 2023 called for a legally binding instrument by 2026 — argue that fully out-of-the-loop systems raise serious problems under international humanitarian law (IHL), particularly the principles of distinction, proportionality, and precaution under Additional Protocol I to the Geneva Conventions. They also raise accountability gaps: if an autonomous system commits an unlawful killing, it is unclear whether the commander, programmer, or manufacturer bears responsibility.

Proponents argue such systems may react faster than humans in saturated environments (e.g., missile defense like the Aegis Combat System or Israel's Iron Dome, which already operate in highly automated modes against incoming projectiles), and could in theory reduce collateral damage by removing emotion and fatigue from targeting decisions.

The distinction matters in MUN and policy drafting because most proposed regulatory frameworks do not seek to ban all autonomy — only out-of-the-loop engagement of humans as targets.