FHI EXPLANATORY REPORT

FHI Explanatory Report.pdf

Download PDF • 192KB

Author's Foreword

My interest in the protection of civilians in armed conflict stems from academic training and long-standing personal commitment. As a law student, I studied under Professor Colonel Gerald Draper, whose contributions to post-war war-crimes jurisprudence shaped modern interpretations of the Geneva Conventions and the principles of distinction and humanity. It was under his influence that I undertook a detailed study of the first Nuremberg Trial, reading the full transcript over two weeks - an exercise that impressed upon me the fundamental purpose of humanitarian law: to protect the innocent.

Two more recent influences crystallised my conviction that the existing framework requires strengthening. The first was Daniel Ellsberg's The Doomsday Machine: Confessions of a Nuclear War Planner (2017), which I read during the COVID-19 lockdown. Ellsberg describes how, in 1961, he drafted a question for President Kennedy to the Joint Chiefs of Staff: "If your plans for general nuclear war are carried out as planned, how many people will be killed in the Soviet Union and China?" The answer - a graph showing 275 to 325 million deaths within six months - was horrifying enough. But when Ellsberg drafted a follow-up question asking for global casualties including fallout in neutral countries, the total rose to six hundred million: Finland wiped out by fallout from attacks on Leningrad, a hundred million dead in India, Afghanistan, Japan, and Western Europe. Ellsberg recalls thinking: "This piece of paper should not exist. It should never have existed. Not in America. Not anywhere, ever."

What struck me most forcibly was not merely the scale of contemplated destruction, but the moral architecture that made such calculations possible. The deaths of hundreds of millions of civilians in neutral countries were categorised as "collateral damage" - foreseeable, accepted, and deemed proportionate to military objectives. Ellsberg traces how this moral architecture developed: from Roosevelt's 1939 appeal to all belligerents to spare civilian populations, unanimously affirmed, to the silent abandonment of that commitment within three years; from the RAF's perfection of firestorm technology at Hamburg and Dresden to the USAAF's systematic incineration of Japanese cities; from the atomic bombings (about which, Truman famously said, he never lost a night's sleep) to the Cold War planning that treated six hundred million deaths as an acceptable cost of "winning."

Reading Ellsberg prompted me to write a two-part analysis, The Inherent Flexibility of the Human Moral Compass, tracing this trajectory from the League of Nations' unanimous 1938 resolution against civilian bombing through to contemporary drone strikes on wedding parties in Afghanistan. The pattern is consistent: once a decision is taken that killing civilians is "acceptable" in pursuit of military objectives, subsequent effort tends to focus on doing so more efficiently. The principle of civilian immunity erodes not through dramatic repudiation but through incremental redefinition of what counts as "excessive" harm.

The second recent influence was the film Eye in the Sky (2015), in which military commanders agonise for over an hour about whether to authorise a drone strike that will kill three high-value terrorist targets when a child selling bread nearby will foreseeably also be killed. The film depicts precisely the kind of deliberation that should occur: verification of civilian presence, consideration of alternatives, escalation to senior authority, genuine moral struggle. Yet the documented frequency of civilian casualties in actual drone operations - wedding parties struck in Haska Meyna, Wech Baghtu, and Musa Qala; rescuers killed in "double-tap" strikes at Mir Ali; pine nut farmers killed at Wazir Tangi - suggests that such deliberation does not, in practice, take place. If it did, we would not have so many foreseeable, avoidable deaths.

One case crystallises the problem. On 21 May 2016, President Obama - not a leader given to cavalier disregard of civilian life - authorised a drone strike against Mullah Akhtar Mohammad Mansour, then leader of the Taliban, as he travelled by taxi through Balochistan. The taxi driver, Mohammad Azam, was a father of four with no connection to any militant group; he had been randomly assigned the fare that morning through a rental car company at the border crossing. The strike killed both men.

What followed was not a proportionality analysis but a definitional evasion. A US official told Reuters that the strike had targeted Mansour and "another combatant." The Pentagon subsequently reported "no civilian casualties." Azam - a commercial driver whose family had no connection to any armed group - was reclassified out of civilian status altogether. This was consistent with a broader Obama-era methodology, documented by the New York Times, that counted "all military-age males in a strike zone as combatants" unless posthumous intelligence proved them innocent. The result was a direct inversion of Article 50(1) of Additional Protocol I, which establishes that in cases of doubt, a person shall be considered a civilian.

Mohammad Azam's death is the paradigm case for the Principle of Foreseeable Harm to Innocents. He was identifiable as a civilian. His death was foreseeable with certainty. It was avoidable - a ground operation, a different time, a different location. On this occasion there was no "ticking bomb." Yet the legal framework failed him not merely through permissive proportionality but through a prior evasion: reclassifying him as a combatant meant that the questions of proportionality and precaution were never reached. If humanitarian law cannot protect Mohammad Azam, it has ceased to perform its protective function.

The Principle of Foreseeable Harm to Innocents is offered as a further-protective interpretation of existing IHL obligations, consistent with the object and purpose of Additional Protocol I. It seeks to close the gap between what humanitarian law promises, and what it delivers; to ensure that when civilians are identifiable, when their deaths are foreseeable, and when harm is avoidable, the attack must not proceed.

The Declaration's intellectual foundations draw on three bodies of scholarship: Jens David Ohlin's defence of the intent/foresight distinction, which demonstrates that foreseeable harm cannot coherently be addressed by reconstructing intent and therefore requires inquiry elsewhere in the doctrinal structure; Neta Crawford's analysis of systemic collateral damage, which exposes the accountability gap that intent-focused frameworks produce; and Henry Shue's argument that legal obligation must be indexed to operational capacity. FHI develops Ohlin's two-track system by adding a logically prior avoidability inquiry that addresses a question his defence does not purport to answer: what constraints apply when civilian harm is not merely foreseen but certain, and when alternatives exist that would avoid it.

I dedicate this work to Gerald Draper, whose example continues to guide my thinking on the law of armed conflict, and to the memory of Mohammad Azam and the countless other foreseeable harm to innocents whose deaths current law has failed to prevent.

Neil Cameron

1. Introduction

This Explanatory Report accompanies the Declaration on Foreseeable Harm to Innocents, a political commitment designed to strengthen civilian protection in contemporary armed conflict. The Declaration articulates the Principle of Foreseeable Harm to Innocents (FHI), which establishes a presumption against attacks where ascertainable civilians face a foreseeable risk of death or serious injury.

The Declaration's aims are twofold: first, to clarify that existing IHL obligations of "constant care" and "feasible precautions" require that foreseeable harm to ascertainable civilians be prevented where alternatives exist; and second, to provide an operationally grounded framework that reflects the capabilities of modern intelligence, surveillance, reconnaissance, and precision targeting.

It is axiomatic that FHI must not and does not make new law. The Declaration articulates what existing IHL already requires when interpreted in light of its protective purpose and modern technological capabilities.

This instrument does not amend the Geneva Conventions or Additional Protocol I. It interprets and strengthens their application.

The Declaration’s structure reflects the sequence inherent in modern targeting: visibility makes presence foreseeable; foreseeability, combined with current capabilities, often makes harm avoidable; and avoidability triggers the legal obligation not to proceed.

2. Purpose of the Declaration

For decades, the rules governing civilian harm - distinction, precaution, proportionality - have been interpreted permissively, allowing militaries to justify predictable civilian deaths, including those captured on live ISR feeds. This permissiveness has contributed to strikes on family vehicles, attacks on residential buildings at night, drone strikes on gatherings, destruction of civilian infrastructure where civilians are known to shelter, and attacks in densely populated urban areas with inadequate verification. Proportionality, as currently interpreted, functions less as a restraint than as a vocabulary of justification.

The Declaration clarifies that existing law already provides a standard grounded in foreseeability rather than intent. Where civilian presence is known or reasonably ascertainable, and where harm is avoidable, the attack must not proceed.

Recent national security doctrines adopted by certain states, including the 2025 United States National Security Strategy, signal an increased willingness to accept higher levels of incidental civilian risk in the name of anticipatory action or strategic competition. These doctrinal shifts heighten the importance of clarifying the legal consequences of foreseeability and avoidability under Article 57 of Additional Protocol I. The Declaration provides such clarity by reaffirming that foreseeable and avoidable harm to civilians remains prohibited regardless of broader strategic or operational posture.

3. Relationship with International Humanitarian Law

A. Compatibility with the Geneva Conventions and Additional Protocol I

The Declaration is fully compatible with existing IHL. It articulates what existing law already requires as a further-protective interpretation of Article 51 (Protection of civilians) and Article 57 (Precautions in attack) of Additional Protocol I, and is anchored in the Martens Clause.

The Declaration does not derogate from or amend existing law. Instead, it provides clarity where IHL is indeterminate, particularly in relation to what counts as "feasible" verification, what qualifies as "excessive" civilian harm, and how ambiguity must be resolved.

This clarification is particularly relevant in an era where some national strategies broaden the scope of operational discretion in the conduct of hostilities. Without interpretive guidance grounded in the protective purpose of Articles 51 and 57, such strategies risk unintentionally diluting obligations that IHL already imposes.

B. Relationship with Proportionality

Proportionality remains part of IHL; the Declaration does not displace it. Rather, the Declaration interprets the proportionality assessment under Article 51(5)(b) to require that foreseeable harm to ascertainable civilians, where avoidable through feasible precautions, is inherently excessive in relation to military advantage. This interpretation is consistent with IHL's protection-favouring principle and gives substantive content to the otherwise indeterminate concept of "excessive" harm.

4. Rationale for the Principle of Foreseeable Harm to Innocents

A. Article 57 Deconstructed: The Textual Foundation of FHI

FHI's claim to be declaratory of existing law rests primarily on Article 57 of Additional Protocol I. The full text of the relevant provisions warrants close examination:

Article 57 – Precautions in attack

In the conduct of military operations, constant care shall be taken to spare the civilian population, civilians and civilian objects.

With respect to attacks, the following precautions shall be taken:

(a) those who plan or decide upon an attack shall:

(i) do everything feasible to verify that the objectives to be attacked are neither civilians nor civilian objects and are not subject to special protection but are military objectives within the meaning of paragraph 2 of Article 52;

(ii) take all feasible precautions in the choice of means and methods of attack with a view to avoiding, and in any event to minimizing, incidental loss of civilian life, injury to civilians and damage to civilian objects;

(iii) refrain from deciding to launch any attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated;

(b) an attack shall be cancelled or suspended if it becomes apparent that the objective is not a military one or is subject to special protection or that the attack may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated;

(c) effective advance warning shall be given of attacks which may affect the civilian population, unless circumstances do not permit.

Each element of this provision supports FHI's interpretive framework:

Constant care (Art. 57(1)). This is not an exhortative aspiration but a binding obligation. "Constant" means continuous - not occasionally, not when convenient, but throughout the conduct of operations. "Care" imports a standard of diligence. FHI operationalises this requirement: where modern ISR makes civilian presence visible, "constant care" demands that visibility be acted upon.

Do everything feasible to verify (Art. 57(2)(a)(i)). The obligation is not to do something feasible, but everything feasible. The scope of "feasible" is capacity-indexed: it depends on the verification means available to the attacker. The 1987 ICRC Commentary confirms that feasibility is measured against "the technical means of identification that are available" and expressly identifies then-emerging reconnaissance technologies - radar, television imaging, infrared sensors, satellites - as part of the legal duty. What was infeasible in 1977 may be routine in 2025. FHI's verification obligations (Article 4 of the Declaration) give operational specificity to this existing requirement.

All feasible precautions... with a view to avoiding, and in any event to minimizing (Art. 57(2)(a)(ii)). Three elements warrant emphasis. First, "all feasible precautions" - not merely reasonable precautions, but all that are feasible. Second, the phrase "with a view to avoiding" establishes avoidance as the primary objective; minimisation ("in any event") is the fallback where avoidance is not achievable. This textual structure directly supports FHI's avoidability gate: the duty is first to avoid, and only where avoidance is infeasible does the analysis shift to minimisation and proportionality. Third, the precautions extend to "choice of means and methods" - encompassing timing, weapon selection, approach vector, and tactical alternatives. FHI's two-stage avoidability test (Articles 5-6 of the Declaration) operationalises what this provision already requires.

Refrain from deciding to launch (Art. 57(2)(a)(iii)). This is a prohibition, not a permission. Attacks that "may be expected" to cause excessive harm shall not be launched. The obligation falls on "those who plan or decide upon an attack" - command responsibility is built into the text. FHI clarifies that where harm is both foreseeable and avoidable, it is inherently excessive: harm that need not occur cannot be justified by proportionality balancing.

Cancelled or suspended if it becomes apparent (Art. 57(2)(b)). The duty of precaution is not discharged at the moment of authorisation; it continues through execution. If circumstances change - if civilian presence becomes apparent - the attack must be cancelled. This dynamic obligation is central to FHI's framework: modern ISR provides real-time information that may require real-time reassessment.

The Capacity-Indexed Structure. Throughout Article 57, "feasible" is the operative qualifier. The ICRC Commentary explains that feasibility depends on "the circumstances ruling at the time, including humanitarian and military considerations" and "the technical means of identification that are available." This structure means that precautionary obligations are not static; they expand as capabilities expand. A force with persistent drone surveillance, pattern-of-life analysis, and precision munitions faces more demanding feasibility requirements than one operating with minimal situational awareness. FHI makes this implication doctrinally explicit through its capacity-indexed framework.

The Logical Priority of Avoidance. The textual sequence of Article 57(2)(a)(ii) is significant: "avoiding, and in any event to minimizing." Avoidance comes first. Minimisation is the residual obligation where avoidance fails. Proportionality analysis under Article 57(2)(a)(iii) presupposes that precautions under 57(2)(a)(ii) have been exhausted. FHI's avoidability gate operationalises this sequence: proportionality is reached only after avoidability has been excluded.

Article 57 thus contains, in compressed form, the doctrinal structure FHI articulates. FHI does not add to these obligations; it unpacks what they already require when applied to the technological conditions of modern warfare.

B. Operational Reality

Modern Intelligence, Surveillance, and Reconnaissance (ISR) capabilities, combined with machine-learning assisted pattern-of-life analysis, give attackers unprecedented visibility of individuals on the ground: children in courtyards, families in vehicles, civilian movement patterns. When the presence of civilians is visible or reasonably inferable, claims of "unknown harm" or "unintended casualties" lose credibility.

C. The Difficulty with the Intent/Foreseeability Distinction

Existing doctrine draws a moral and legal distinction between intended harm and foreseen but unintended harm. This distinction becomes difficult to sustain when operators can see civilians on screens and when collateral damage estimates predict their deaths. The Declaration recognises that certainty of outcome is morally more salient than the attacker's characterisation of mental state. When harm is both foreseeable and avoidable, continued reliance on intention alone ceases to perform any civilian-protective function.

When harm is both foreseeable and avoidable, continued reliance on intention alone ceases to perform any civilian-protective function.

D. The Indeterminacy of Proportionality in Practice

The proportionality test has become inconsistent in application, susceptible to manipulation, opaque to external review, and detached from the experience of affected civilians. FHI addresses this by specifying that where civilian presence is foreseeable and harm is avoidable, that harm is excessive as a matter of law.

E. The Imperative of Risk Reallocation

A fundamental principle of just war tradition, reflected in modern human rights jurisprudence, holds that combatants must accept risk rather than transfer it to civilians. As Michael Walzer has argued, "if saving civilian lives means risking soldiers' lives, the risk must be accepted." The Declaration incorporates this principle directly: where precautions can reduce civilian harm by increasing risk to the attacking force, such reallocation shall be undertaken where feasible and where the reduction in civilian harm is significant.

Contemporary strategic doctrines that emphasise force protection, anticipatory action, or rapid operational tempo can create institutional pressures that favour accepting higher levels of civilian risk. The Declaration responds to these pressures not by introducing new obligations, but by clarifying that existing precautionary duties require preventing foreseeable and avoidable harm to civilians even when operational incentives might point in the opposite direction.

F. The Special Protection of Children

Children bear a disproportionate share of foreseeable harm in contemporary conflict. Their visibility, predictability of presence (in homes, schools, and with families), and particular vulnerability justify explicit attention under the Declaration.

G. Capacity-Indexed Obligations: A Doctrinally Latent Requirement

Article 57’s feasibility standard is intrinsically capacity-indexed: the more an attacker can know, the more an attacker must do. The travaux préparatoires and ICRC Commentary explicitly tie precautionary duties to the “technical means available at the time.” FHI makes this implicit principle explicit. Modern militaries possess ISR capabilities far beyond those envisioned in 1977, and their duties to foresee and avoid civilian harm correspondingly expand. Capacity-indexed obligations therefore form a structural foundation for the avoidability predicate.

Under the Principle of Foreseeable Harm to Innocents, legal obligation thickens only as capacity thickens: where actors lack the practical ability to know or to do otherwise, moral tragedy may remain, but legal fault does not arise.

5. Structure of the Declaration

The Declaration consists of nineteen articles organised into eight Parts, plus an Interpretive Guidance Annex:

• Part I: General Provisions (Articles 1–2): Purpose, scope, application, and definitions

• Part II: Substantive Obligations (Articles 3–7): The core FHI principle, duties to verify, minimise harm, reallocate risk, the prohibition on foreseeable avoidable harm and the use of human shields

• Part III: Emergency Override (Article 8): The narrow exception for genuine mass-casualty emergencies

• Part IV: Procedural Obligations (Articles 9–11): Documentation, transparency, post-incident review, and accountability

• Part V: Remedies for Victims (Article 12): Reparation and accessible procedures

• Part VI: Relationship with International Law (Article 13): Compatibility with IHL

• Part VII: Implementation (Articles 14–16): Domestic measures, non-state armed groups, international cooperation

• Part VIII: Norm Development and Final Provisions (Articles 17–19): Judicial consideration, customary law formation, endorsement and review

6. The FHI Decision Sequence in Operation

Before examining each Article in detail, it is useful to see how FHI operates as a sequential decision framework. The stages must be applied in order; each is a gate that must be passed before proceeding to the next.

Stage 1 - Status Determination. Is the person directly participating in hostilities? If yes, FHI does not apply and standard IHL targeting rules govern. If no, or if doubt exists, that doubt is resolved in favour of civilian status and the analysis proceeds.

Stage 2 - Verification. Has civilian presence been verified through all feasible measures? Verification duties are capacity-indexed: forces with greater ISR capability bear correspondingly greater obligations.

Stage 3 - Foreseeability Assessment. Are persons identifiable as civilians foreseeably at risk? If no, the attack may proceed subject to standard IHL. If yes, the attack is presumptively prohibited and the analysis continues.

Stage 4 - Avoidability Gate. This is FHI's central contribution. It comprises two tests applied in sequence:

Stage 4A - Elimination Test: Can any feasible alternative - temporal, tactical, positional, operational, or non-kinetic - eliminate the foreseeable harm? If yes, the attack is prohibited and the alternative must be employed.

Stage 4B - Substantial Reduction Test: If elimination is not feasible, can any alternative substantially reduce the foreseeable harm? If yes, the attack is prohibited unless the harm-reducing alternative is adopted.

Only if both tests are satisfied - harm cannot be eliminated or substantially reduced - is the harm "genuinely unavoidable" and the analysis proceeds.

Stage 5 - Risk Allocation. Have combatants accepted risk rather than transferring it to civilians? If combatants could reduce civilian risk by accepting additional exposure but have not done so, the attack is prohibited.

Stage 6 - Proportionality. Is expected civilian harm excessive in relation to concrete and direct military advantage? Proportionality analysis applies only at this stage — only to the residual category of genuinely unavoidable harm.

Stage 7 - Emergency Override. In exceptional circumstances where Stages 4–6 cannot be satisfied, but catastrophic imminent harm threatens, the eight cumulative conditions of Article 8 may permit the attack. This override is expected to apply rarely, if ever.

The sequence embeds two principles:

First, avoidability precedes proportionality. The question "could this harm have been avoided?" must be answered before "is this harm proportionate?" Foreseeable and avoidable harm is inherently excessive.

Second, the presumption favours restraint. Under FHI, protection is the default when ascertainable civilians are foreseeably at risk. That presumption may be overcome only through the staged analysis set out above.

7. Explanation of the Articles

Article 1: Purpose, Scope, and Application

Article 1 clarifies five issues:

Complementary character: The Declaration supplements existing law and neither replaces nor weakens Geneva rules, shielding it from objections that it fragments the law.

Applicability to all armed conflicts: The explicit extension to non-international armed conflicts prevents states from limiting the Declaration's applicability by reclassifying operations; most modern conflicts are NIACs and the Declaration must cover them.

Extraterritorial application: The Declaration explicitly applies to all attacks "whether conducted within or outside the territory of an endorsing entity, including drone strikes, airstrikes, artillery fire, special operations, and any other use of force." This ensures FHI covers the full range of modern military operations.

Non-derogation: Nothing in the Declaration diminishes existing IHL or human rights obligations.

Promotion: Endorsing entities commit to promote these standards to all parties, including non-state armed groups.

Article 2: Definitions

Article 2 provides ten definitions that form the analytical backbone of the Declaration:

Civilians whose presence makes harm foreseeable identifies civilians whose presence is known or reasonably ascertainable through available means. This operationalises the reality that civilians are visible, their presence is predictable, and ISR capabilities detect them.

Foreseeable risk establishes the threshold: reasonable grounds to believe death or serious injury will result.

Avoidable harm integrates the alternative-measures doctrine - harm that can be prevented through feasible alternatives.

Unlawful risk transfer codifies the principle that combatants may not shift risk to civilians when they could bear it themselves.

Feasible adopts the established IHL standard: practicable under the circumstances.

Feasible alternatives elevates alternatives from a mere consideration to a determinative requirement.

Attack uses the standard IHL definition while excluding cyber operations not expected to cause physical harm.

Military advantage is defined restrictively: concrete and direct, excluding speculative benefits.

Commander covers all persons with targeting authority.

Emergency override is defined narrowly: imminent mass-casualty events only.

Article 3: The Principle of Foreseeable Harm to Innocents

Article 3 is the central provision. It establishes:

The core commitment (paragraph 1): Attacks shall not be launched where ascertainable civilians face a foreseeable risk of death or serious injury, except where the conditions of Article 6 are satisfied. This formulation focuses on outcome rather than intent: what matters is not the attacker's state of mind but whether ascertainable civilians are foreseeably at risk.

The direct participation exclusion (paragraph 2): The protections of Article 3 do not extend to persons directly participating in hostilities for such time as they so participate. This language tracks Article 51(3) of Additional Protocol I verbatim, ensuring doctrinal continuity. The exclusion operates as a threshold matter: a person who is directly participating in hostilities is not a "civilian" entitled to the presumption against attack, and the FHI framework does not apply to targeting decisions concerning such persons.

The phrase "for such time as" preserves the temporal limitation recognised in IHL: direct participation suspends but does not permanently forfeit civilian protection. A civilian who takes a direct part in hostilities may be targeted during that participation, but regains protection when the participation ceases.

The Declaration deliberately refrains from elaborating the criteria for direct participation, instead deferring to "applicable international humanitarian law." This reflects the ongoing doctrinal debate surrounding the ICRC's 2009 Interpretive Guidance on the Notion of Direct Participation in Hostilities, which some states and commentators have criticised as either too restrictive or too permissive. FHI does not seek to resolve that debate; it incorporates whatever standard of direct participation is applicable under existing law. This ensures that FHI strengthens protection for civilians without becoming entangled in contested questions about the boundaries of DPH.

Importantly, the DPH exclusion does not create a loophole. Where doubt exists as to whether a person is directly participating in hostilities, the presumption of civilian status under Article 4(3) applies. The exclusion operates only where direct participation is established, not where it is merely suspected.

The presumption against attack (paragraph 3): When ascertainable civilians whose presence makes harm foreseeable are present or likely to be present, the default position is that the attack must not proceed. This inverts the operative logic of permissive proportionality interpretations, which tend to treat attacks as presumptively lawful unless harm is deemed "excessive." Under FHI, the presumption favours protection; the burden lies with the attacker to demonstrate that the narrow conditions of the emergency override are satisfied.

Anchoring in Additional Protocol I (paragraph 4): FHI is not a novel legal creation but a clarification of what Articles 57(1) and 57(2) of Additional Protocol I already require. Those provisions impose duties of "constant care" and "all feasible precautions" to spare civilians. FHI gives these duties operational content: where civilian presence is foreseeable and harm is avoidable through the two-stage avoidability test, the precautionary obligation is not merely to "consider" alternatives but to employ them — including, where necessary, cancellation of the attack.

Relationship with proportionality (paragraph 5): The proportionality assessment under Article 51(5)(b) remains part of IHL; FHI does not displace it. However, FHI interprets proportionality to specify that foreseeable harm to ascertainable civilians, where avoidable, is "inherently excessive" in relation to military advantage. This interpretation follows from the logic of the avoidability gate: if feasible alternatives exist that would achieve the military objective while sparing civilians, then proceeding with an attack that kills those civilians cannot be proportionate. Proportionality analysis proper - the weighing of expected civilian harm against anticipated military advantage - arises only in the residual category of cases where harm is foreseeable but genuinely unavoidable.

Article 4: Duty to Verify Civilian Presence

Article 4 operationalises verification:

All feasible measures: Commanders must use available verification capabilities.

Specified measures: Intelligence analysis, real-time surveillance, pattern-of-life assessment, collateral damage estimation, local knowledge consultation, and sensor platforms.

Doubt resolved for civilians: Both as to status and as to presence.

Failure as violation: Not verifying when verification was feasible is itself a breach.

The avoidability inquiry proceeds through a two-stage test. First, whether any feasible alternative could eliminate foreseeable harm. Second, whether feasible alternatives could substantially reduce that harm. Only where both stages yield negative results is harm considered unavoidable. This interpretation aligns with, and gives practical effect to, Article 57’s capacity-indexed precautionary structure.

The US DoD Law of War Manual (2023 revision) includes a dedicated section on 'Feasible Precautions to Verify Whether the Objects of Attack Are Military Objectives' (§ 5.5.3), demonstrating that FHI's verification requirements are operationally achievable. However, the Manual's progressive language coexists with permissive targeting practices; the gap between doctrinal commitment and operational reality is precisely what FHI seeks to close. FHI builds on these foundations by providing the binding force that doctrine alone has lacked.

Article 5: Duty to Minimise Harm and Reallocate Risk

Article 5 addresses precautions and risk:

All feasible precautions: Timing, weapons, approach, postponement, cancellation, non-kinetic means, capture operations.

Risk reallocation: Where precautions reduce civilian harm by increasing combatant risk, reallocation is required where feasible and where reduction is significant. This reflects the established principle that combatants must accept risk rather than transfer it to civilians.

No excuse: Operational convenience, efficiency, cost, or force protection cannot justify failure to take feasible precautions.

Article 6: Prohibition Where Harm Is Foreseeable and Avoidable

Article 6 establishes the operative commitment:

Core rule: Endorsing entities will not proceed with attacks where harm is foreseeable and avoidable.

Intent not determinative: "The attacker's subjective intent does not determine the lawfulness of an attack under this Declaration. Foreseeability of harm, not intent to cause it, is the operative criterion."

When presence is foreseeable: Visible on surveillance; residential structures at night; civilian vehicles; displaced persons shelters; pattern-of-life indicators; densely populated areas; contextual indicators.

Grave violation: Breaches may trigger criminal liability under Article 11.

Article 7: Human Shields

Article 7 addresses the complex situation where civilians are used as shields:

 (a) Voluntary shields. Civilians who voluntarily and directly shield military objectives from attack may be considered directly participating in hostilities for such time as they do so. They do not qualify as "persons identifiable as civilians" for purposes of the FHI presumption, though the attacking party retains an obligation to verify their status and intent before treating them as participants.

(b) Involuntary shields. Civilians forced, coerced, or prevented from leaving the vicinity of military objectives retain full civilian status. The shielding party's unlawful conduct does not extinguish the attacking party's precautionary obligations.

(c) Responsibility allocation. Where an attack proceeds after exhausting feasible precautions, moral and legal responsibility may be allocated primarily to the shielding party rather than the attacking party, provided the attacker's conduct satisfied FHI's procedural requirements.

(d) Limits. The presence of human shields does not suspend proportionality. An attack causing civilian harm grossly disproportionate to the military advantage anticipated remains unlawful regardless of how the civilians came to be present.

Article 8: Emergency Override

Article 8 provides the safety valve - but a narrow one:

Eight cumulative conditions:

1. Imminence: A specific mass-casualty attack is imminent (minutes or hours, not days), threatening tens, hundreds, or thousands of civilian deaths. Speculative or temporally remote harms do not qualify.

2. No feasible alternative: No tactical, temporal, positional, operational, or non-kinetic alternative exists; the strike is the only means of averting the mass-casualty event. All alternatives must be documented and rejected.

3. Unavoidability confirmed: Harm to persons identifiable as civilians is genuinely unavoidable as determined through the two-stage avoidability test—not merely inconvenient to avoid.

4. Decisive asymmetry: Expected casualties from the strike are substantially lower—by a very large margin—than casualties prevented. The harm prevented must vastly exceed the harm caused.

5. Irreversibility: The threatened harm cannot be mitigated by subsequent action. If later intervention could avert or reduce harm, immediate action is not justified.

6. Intelligence reliability: Intelligence is current, multi-source where feasible, and highly reliable.

7. Highest-level authorisation: Authorised by a pre-designated authority at the highest level practicable.

8. Contemporaneous documentation and automatic review: The commander records, at the time of decision, the factual basis for invocation—including intelligence, alternatives rejected, and asymmetry calculation. Independent post-strike review is automatically triggered.

Invocation presupposes that sufficiently reliable intelligence was available to the decision-maker at the time of attack.

Exclusions: High-value targets (unless actively directing imminent attack); strategic degradation; battlefield advantage; political or morale considerations; speculative harm; targets of opportunity.

Expected rarity: "This provision is expected to apply rarely, if ever. Frequent invocation... indicates either misapplication of its conditions or circumstances requiring doctrinal reassessment."

Review still required: Invocation does not exempt from subsequent review.

Article 9: Documentation and Evidentiary Burdens

Article 9 makes accountability possible:

Burden on attacker: The attacking party must demonstrate compliance.

Comprehensive retention: Surveillance recordings, targeting records, CDEs, alternatives assessments, risk-allocation determinations, legal advice, intelligence summaries.

Ten-year retention: Ensuring long-term accountability.

Redaction permitted: But essential compliance information must be preserved.

Adverse inference: Missing documentation creates presumption against attacker.

Tampering presumption: Destruction or manipulation creates presumption of unlawful conduct.

Article 10: Transparency and Post-Incident Review

Article 10 requires:

Prompt, impartial, effective review of every civilian harm incident.

Public findings to extent consistent with operational security.

Transmission to oversight bodies and affected communities.

Corrective measures where violations found.

Aggregate reporting: Periodic public data on incidents, investigations, and accountability.

Article 11: Accountability

Article 11 establishes two-tier accountability:

Tier 1: Grave Violations (Criminal Responsibility):

• Attacks in clear violation where civilian presence was known

• Knowingly misrepresenting intelligence

• Reckless disregard amounting to wilful blindness

• False invocation or abuse of emergency override

• Falsifying or destroying documentation

• Command responsibility failures

These may constitute war crimes and must be incorporated into national criminal law. Criminal procedural protections are preserved.

Tier 2: Serious Violations (Administrative Responsibility):

• Failure to verify

• Failure to take precautions

• Failure to escalate

• Failure to document

Administrative review is triggered; operator may be required to demonstrate due diligence. Consequences include reprimand, retraining, suspension, loss of promotion, operational restrictions, or separation.

Article 12: Reparation for Victims

Article 12 addresses those harmed:

Effective remedy: Acknowledgment, compensation, medical/rehabilitative assistance, truth recovery, non-repetition assurances.

Accessible procedures: Including for persons in conflict-affected areas.

Condolence payments: Not contingent on violation finding; not admission of liability; but do not discharge full reparation obligation where violation established.

Article 13: Relationship with Existing International Humanitarian Law

Article 13 affirms:

Further-protective interpretation: The Declaration articulates what existing law already requires; it does not amend it.

Consistency: With Geneva Conventions and Martens Clause.

More protective prevails: Where Declaration and other IHL address same matter.

No authorisation of violations: Nothing permits otherwise unlawful conduct.

Article 14: Domestic Implementation

Article 14 requires:

Incorporation: Into legislation, manuals, doctrine, ROE, training.

Legal adviser training: Judge advocates must be trained in FHI.

Advance designation: Of emergency override authority.

Coalition operations: FHI applies where endorsing states contribute forces.

Implementation Guidelines: States are encouraged to consult the Model Implementation Guidelines adopted alongside the Declaration.

Article 15: Engagement with Non-State Armed Groups

Article 15 addresses NSAGs:

All parties bound: By IHL, encouraged to respect FHI.

Support for promotion: Via ICRC, UN, others.

Unilateral declarations: Welcomed; may be deposited with UN Secretary-General.

Status unaffected: Endorsement does not affect legal status.

Article 16: International Cooperation

Article 16 commits to:

Sharing: Best practices, joint training, technical assistance.

Monitoring: Support for international mechanisms.

Coordination: Of investigations involving multiple parties.

Promotion: In UN, regional organisations, IHL processes.

Article 17: Judicial Consideration

Article 17 provides:

Evidentiary value: Courts may have regard to Declaration as evidence of state understanding.

Declaratory: Does not bind courts.

Encouragement: Litigants encouraged to cite FHI in proceedings.

Article 18: Norm Development

Article 18 anticipates:

Progressive development: Contribution to IHL development and customary law.

Promotion: Through state practice and official statements.

Future treaty: Not precluded; willingness expressed.

Article 19: Endorsement, Review, and Amendment

Article 19 establishes:

Open endorsement: States, international organisations, civil society.

Depositary: UN Secretary-General; public register.

Review conferences: Within three years, then periodic.

Reporting: Periodic implementation reports.

Amendment: By consensus; enters into force for accepting entities.

8. The Interpretive Guidance Annex

The Declaration includes an Annex providing guidance on key interpretive questions. This Annex does not form part of the operative text but assists implementation:

• Persons identifiable as civilians whose presence makes harm foreseeable: Reasonableness standard; capability-dependent obligations

• Feasible: Not to be conflated with convenience

• Presumption: Not rebutted by mere existence of military objective

• Risk reallocation: Does not require suicidal operations; does require acceptance of increased risk where it significantly reduces civilian harm

• Emergency override: Frequent invocation indicates misapplication

• Two-tier accountability: Rationale for criminal/administrative distinction

• Victim remedies: Condolence payments distinguished from reparation

9. Relationship to Contemporary Conflicts

The Declaration is particularly responsive to patterns documented in Gaza, Ukraine, Yemen, Somalia, Afghanistan, and Syria. In these conflicts, foreseeable civilian deaths have been justified under existing doctrinal interpretations. FHI addresses these gaps by clarifying that where civilian presence is foreseeable and harm is avoidable, existing law already prohibits the attack.

10. Anticipated Objections and Responses

"This eliminates proportionality."

No. Proportionality remains applicable. FHI interprets proportionality consistent with the principle of humanity, specifying that foreseeable, avoidable harm is inherently excessive.

"This weakens military necessity."

Military necessity has never justified foreseeable killing of civilians where alternatives exist. FHI makes explicit what the principle of necessity already implies.

"We already comply with IHL."

Evidence from conflict zones, documented by UN bodies, NGOs, and investigative journalists, contradicts claims of universal compliance. FHI provides a clearer standard against which compliance can be measured.

"Intent matters."

Intent remains relevant to moral and legal assessment. But when harm is foreseeable with near-certainty, the moral significance of "not intending" that harm diminishes. FHI shifts the focus to what the attacker knew would happen, not what the attacker claims to have hoped.

"This is too strict."

Civilian protection is meant to impose genuine constraints; flexibility already lies overwhelmingly with militaries. FHI restores balance.

"This is unrealistic."

Modern ISR and precision capabilities make FHI entirely feasible. The Declaration requires only what technology already enables.

"This criminalises mistakes."

Only reckless, negligent, or deliberate failures attract criminal liability under Tier 1. The administrative tier (Tier 2) handles procedural failures without requiring proof of criminal intent.

11. Pathways to Adoption

The Declaration is structured to facilitate:

• Endorsement by states and NGOs

• Dissemination via UN agencies

• Inclusion in military doctrine and rules of engagement

• Judicial recognition as evidence of state understanding of IHL

• Eventual transformation into a binding treaty instrument or additional protocol

12. Conclusion

The Declaration on Foreseeable Harm to Innocents represents a principled, operationally realistic, and urgently needed clarification of existing humanitarian law. It is rooted in established jurisprudence, aligned with the Geneva Conventions, and motivated by the contemporary realities of conflict.

It aims to ensure that identifiable innocents are protected from foreseeable harm - restoring credibility and humanity to the conduct of hostilities, and giving effect to the promise that civilians shall be spared the violence of war.