Legal and Ethical Considerations for Coalition Operations in Proximity to Autonomous Weapons Systems: Kaltovia Operational Context

1. Purpose and Status

This advisory sets out the principal legal and ethical considerations arising from coalition operations in an environment that includes an adversary autonomous area-denial system of the type described in TEC-2025-0998. It is intended to inform planning decisions and to ensure that relevant legal frameworks are visible to operational planners at the appropriate stage of the planning cycle. This document is an advisory product. It does not constitute legal clearance for any specific operation, nor does it substitute for the case-specific legal review that any proposed operation requires prior to execution. Legal frameworks addressed here are living bodies of doctrine subject to ongoing interpretive development.

2. Applicable Legal Frameworks

Operations in this planning context are governed by the law of armed conflict as codified in the Geneva Conventions (1949) and Additional Protocols (1977) and in customary international humanitarian law. The relevant principles are: distinction (parties must distinguish between combatants and civilians and may only direct attacks against military objectives — AP I, Art. 48 and Art. 52); proportionality (anticipated civilian harm must not be excessive relative to expected military advantage — AP I, Art. 51(5)(b)); precaution (all feasible precautions must be taken to avoid or minimise civilian casualties — AP I, Art. 57); and necessity (force may only be used to the extent required to achieve a legitimate military objective — customary norm, reflected in AP I, Art. 35). These principles apply with particular acuity where operations involve or foreseeably trigger an autonomous system with an assessed civilian misidentification rate.

The concept of meaningful human control over lethal force is addressed across multiple instruments and processes including the REAIM Declaration (2022, 2024), the UN Group of Governmental Experts on LAWS (2014–present), and the Political Declaration on Responsible Military Use of AI (2023). While these instruments do not create binding treaty obligations in their current form, they represent the emerging international consensus and inform the interpretive environment within which coalition operations are assessed. The common thread: decisions to use lethal force should not be wholly delegated to automated systems where human review of individual targets is feasible. A system designed to engage without operator review of individual engagements — as Site-7 is assessed to be — raises compliance questions that cannot be resolved by reference to the adversary’s technical choices alone.

The assessed 23% civilian misidentification rate of Site-7 (TEC-2025-0998), as a foreseeable consequence of operations that trigger the system’s autonomous engagement authority, is a factor that planners must incorporate into proportionality and precaution assessments. The fact that the system is adversary-operated does not extinguish coalition legal obligations regarding foreseeable consequences of coalition actions. Conversely, an operation that successfully neutralises Site-7 before activation — thereby preventing autonomous engagements — may be assessed as reducing civilian harm risk, provided the operation itself complies with IHL principles. This is a contextual assessment that depends on the specifics of the planned operation.

3. Planning Implications

CLAD recommends that the following considerations be integrated into the planning process: approach axis selection should account for the proportionality implications of triggering Site-7’s coverage along any axis that places civilian populations within the assessed zone; force composition and rules of engagement should reflect precautionary obligations applicable to operations near an autonomous system with documented misidentification characteristics; and the proposed operational timeline — insofar as it is shaped by the 5 November activation deadline — reflects a genuine legal rationale, as the pre-activation window is the window in which human control over engagement decisions can be preserved.

Legal review of the specific operational plan, once developed, should be requested from CLAD at the earliest practicable stage. This advisory is not a substitute for that review. For contextual academic depth on meaningful human control and autonomous weapons in authoritarian state deployment contexts, see ACA-2024-0089.

4. Targeting Review and Rules of Engagement Considerations

Any operational plan submitted for CLAD legal review should include a targeting review in conformity with the coalition targeting directive (classified reference withheld). For operations of the type under consideration, the targeting review must address: (a) the legal basis for the operation under the authorising framework; (b) the identification and classification of all anticipated targets — including the Site-7 installation itself and any associated infrastructure; (c) the proportionality assessment for each planned strike or engagement, accounting for the foreseeable consequences of triggering Site-7’s dead-hand protocol; and (d) the precautionary measures incorporated into the plan to reduce civilian exposure.

Rules of engagement must address the specific scenario in which coalition forces operate in proximity to an adversary autonomous system that may engage without warning or human authorisation. The standard escalation-of-force procedures that apply to defended installations are not fully applicable where the adversary system does not require operator decision at the point of engagement. ROE should be designed to ensure that coalition forces do not inadvertently provide the activation triggers — movement, thermal signatures, electromagnetic emissions — that the system uses for target identification. Specific ROE guidance on proximity to autonomous systems is in preparation; planners should consult CLAD’s IHL and Operational Compliance Branch directly for interim guidance.

Coalition member states have differing positions on the permissibility of operations involving or foreseeably triggering adversary autonomous weapons systems. Two coalition partners have submitted reservations to the REAIM Declaration’s provisions on precaution in LAWS environments. This advisory reflects the consensus framework and planners should be aware that member-state-specific positions may apply to national contingent ROE. CLAD can provide member-state position summaries on request.