FHI Policy Brief

FHI Policy Brief.pdf

Download PDF • 196KB

1. What the Declaration Does

The Declaration on Foreseeable Harm to Innocents establishes a clear and operational civilian-protection rule:

Attacks shall not be launched where persons clearly identifiable as civilians are foreseeably at risk of death or serious injury, unless narrowly defined emergency conditions are met and all feasible alternatives have been exhausted.

This rule includes an explicit avoidability gate. If harm to ascertainable civilians is both foreseeable and avoidable - given the attacker's actual capabilities - the attack must not proceed. Only when harm is foreseeable but genuinely unavoidable, even after all feasible precautions, does the operation move to proportionality analysis.

Proportionality analysis arises only after avoidability has been exhausted,

The Declaration also incorporates capacity-indexed obligations: forces with greater ISR, precision, and analytical capacity bear correspondingly thicker legal duties to prevent civilian harm. This preserves equity between actors while aligning legal obligations with technological reality.

The Principle of Foreseeable Harm to Innocents (FHI) thus moves IHL toward an outcome-based standard supplementing subjective intent-based reasoning where harm is unforeseeable, but displaces it where civilian harm is both foreseeable and avoidable with a clear, evidence-grounded framework that conforms fully to existing treaty law and the protective logic of Article 57 of Additional Protocol I.

The Declaration's logic rests on a simple sequence: modern ISR makes civilian presence visible; visibility makes harm foreseeable; foreseeability, combined with current capabilities, makes much civilian harm avoidable; and avoidability triggers a legal obligation not to proceed.

The concept of capacity-indexed obligations is central: the scope of the precautionary duty scales with the attacker's capability. What was infeasible in 1977 may be routine in 2025.

Foreseeable and avoidable harm to innocents is impermissible harm.

2. Why This Is Needed Now

Civilian deaths are increasingly foreseeable, and increasingly avoidable

In recent conflicts from Gaza to Ukraine, Yemen to Somalia, civilians (especially children) have been killed in circumstances where their presence was visible on sensors, inferable from context, or detectable through standard ISR processes. These casualties can no longer be described as 'random' or 'accidental': they are the predictable result of decisions made despite the existence of feasible alternatives.

The intent/foresight distinction has collapsed

International tribunals have tied themselves in knots trying to define 'intent.' As Dean Jens David Ohlin of Cornell Law School documented in his influential article Targeting and the Concept of Intent, the ICTY's Galić jurisprudence effectively collapsed the two tracks of IHL - distinction and proportionality - by importing dolus eventualis into the concept of intent. If foreseeing civilian casualties counts as 'intending' to attack civilians, proportionality analysis becomes superfluous. Yet the law clearly contemplates that commanders will foresee collateral damage and proceed anyway, provided the harm is not excessive.

FHI resolves this paradox. Rather than collapsing the two-track system, FHI inserts an avoidability inquiry before proportionality - preserving the distinction between targeting civilians and foreseeing their deaths while closing the gap that distinction leaves open: the case where harm is foreseeable, probable, and avoidable.

Existing interpretations are too permissive

The proportionality test has become so elastic that it functions less as a restraint than as a vocabulary of justification, allowing foreseeable civilian deaths to be justified on the basis of subjective claims about military advantage. The reliance on intent - rather than knowledge and capacity - creates a doctrinal gap that shields foreseeable, preventable harm from legal consequence.

As Neta Crawford has documented, most civilian deaths arise not from error but from predictable institutional choices that remain legally insulated - what she terms 'systemic collateral damage.' Existing doctrine provides no framework for addressing harm that is foreseeable, predictable, and the result of deliberate policy choices. FHI fills this gap.

Process reform alone is insufficient

States have responded to civilian harm concerns primarily through institutional reform. The US Department of Defense's Civilian Harm Mitigation and Response Action Plan (CHMR-AP, 2022) created over 160 new positions, established a Civilian Protection Center of Excellence, and standardised reporting procedures. Yet independent assessments measure implementation progress - positions filled, institutions created - rather than reduced civilian harm. Process reform refines how proportionality is applied; it does not change what proportionality does.

The gap between ethical aspiration and legal obligation is not merely theoretical. In May 2013, President Obama declared that 'before any strike is taken, there must be near-certainty that no civilians will be killed or injured - the highest standard we can set.' Yet this standard was adopted as policy rather than legal requirement, and applied only outside 'areas of active hostilities.' Military lawyers advised that existing law did not mandate such stringency, and the standard was accordingly treated as discretionary - capable of being waived for individual operations and ultimately revoked by a successor administration.

FHI seeks to close this gap: not by inventing new obligations, but by demonstrating that the protective standard Obama articulated as aspiration is, properly understood, what the existing legal framework already requires when modern ISR capabilities render civilian harm both foreseeable and avoidable.

Modern technology demands a higher standard

Contemporary targeting environments include real-time ISR, pattern-of-life analysis, data-fusion tools, AI-assisted classification, and precision munitions. Because civilian presence is now often knowable, and because alternatives frequently exist, a rule that treats predictable deaths as 'unfortunate but lawful' is no longer viable.

Combatants must accept risk rather than transfer it to civilians

FHI incorporates the longstanding principle articulated by Michael Walzer and reflected in modern human-rights jurisprudence: 'if saving civilian lives means risking soldiers' lives, the risk must be accepted.' When saving civilian lives requires accepting operational risk, the burden rests with combatants - not with children in residential buildings.

Capacity-disparities demand a capacity-indexed standard

States and armed groups vary enormously in capability. The Declaration preserves fairness by ensuring that obligations scale with actual capacity. A force with advanced ISR cannot justify harm on the basis of the standards applicable to one operating with minimal situational awareness.

FHI therefore clarifies the legal consequences of modern technological reality, rather than creating new obligations.

3. Key Elements of the Declaration

FHI operates as a sequential framework—each element below represents a stage that must be satisfied before proceeding to the next. The sequence is designed so that avoidability is exhausted before proportionality, and the presumption throughout favours civilian protection.

A. Presumption Against Attack

Where persons identifiable as civilians are foreseeably at risk, existing IHL, as clarified by FHI, establishes a presumption against attack. Foreseeable and avoidable civilian harm is inherently excessive - no weighing of military advantage can justify harm that need not occur.

Direct participation exclusion: FHI does not protect persons who are directly participating in hostilities for such time as they so participate. Such persons may be targeted under standard IHL rules. Where doubt exists as to whether a person is directly participating, that doubt is resolved in favour of civilian status.

The two-stage avoidability gate: The presumption against attack may be overcome only where harm is genuinely unavoidable - that is, where all feasible alternatives have been exhausted through the two-stage avoidability test:

Stage 1 - Elimination: Can any feasible alternative eliminate the foreseeable harm entirely? If yes, the attack must not proceed as planned.

Stage 2 - Substantial Reduction: If elimination is impossible, can any feasible alternative substantially reduce the foreseeable harm? If yes, that alternative must be adopted.

Only if both stages are exhausted - only if harm is genuinely unavoidable - does proportionality analysis proceed.

An attack that foreseeably endangers civilians is prohibited unless:

(1) The person is directly participating in hostilities - in which case FHI does not apply; or

(2) All of the following conditions are satisfied:

• Civilian presence is foreseeable

• All feasible alternatives were analysed through the two-stage avoidability test

• Harm is genuinely unavoidable—civilians cannot be spared without negating the legitimate military purpose

• No unlawful risk transfer occurred

• Either proportionality is satisfied, or emergency override criteria are met in full

Where doubt exists as to civilian status, that doubt is resolved in favour of protection.

Proportionality applies only to the residual category where harm is foreseeable but genuinely unavoidable. Foreseeable, avoidable harm is inherently excessive and never reaches proportionality analysis.

B. Duty to Verify

All feasible verification measures must be taken: intelligence analysis, surveillance, pattern-of-life assessment, collateral damage estimation, local knowledge consultation, sensor platforms, review of alternative timings, methods, or angles. Under the capacity-indexed model, forces with better ISR have thicker legal duties to verify. Doubt must be resolved in favour of civilian protection. Failure to verify is itself a violation.

C. Duty to Minimise Harm and Reallocate Risk

If civilians are present, or likely, attackers must take all feasible precautions - including bearing increased operational risk where this significantly reduces civilian harm. Operational convenience, speed, cost, or force protection cannot excuse failure to take feasible precautions.

D. Knowledge, Not Intent, Determines Lawfulness

The attacker's subjective intent is no longer decisive. What matters is: (1) what the attacker knew or should have known, and (2) what they were capable of doing to avoid harm.

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

E. Relationship with Proportionality

FHI does not replace proportionality; it clarifies it. Under the avoidability gate: if harm is avoidable, it is legally excessive - and therefore unlawful - because feasible alternatives exist. This is what existing IHL already requires; FHI does not create new prohibitions but articulates what Article 57's precautionary obligations already require. Proportionality applies only to the narrow set of cases where harm is foreseeable and unavoidable.

Note on civilian objects: FHI's primary focus is protection of civilian lives. Damage to civilian objects remains governed by conventional proportionality standards under Article 52 of Additional Protocol I. However, the avoidability principle applies where destruction of civilian objects would foreseeably cause civilian deaths or serious injury.

F. Human Shields (Article 7)

The Declaration addresses human shields through a responsibility allocation framework:

Voluntary shields are directly participating in hostilities and do not qualify for civilian protection, though verification of status remains required.

Involuntary shields retain full civilian status. The avoidability analysis accounts for the defender's unlawful conduct, feasible precautions taken, alternatives that avoid rewarding shielding, and the time already afforded.

Responsibility allocation: Where an attack proceeds after exhausting feasible precautions, primary responsibility for civilian harm lies with the shielding party. Proportionality remains applicable; grossly disproportionate harm is unlawful regardless of how civilians came to be present.

G. Emergency Override (Article 8)

The override applies only when all eight cumulative conditions are satisfied:

1. Imminence - a specific mass-casualty attack (tens, hundreds, or thousands of deaths) is imminent within minutes or hours, not days

2. No feasible alternative - the strike is the only means of averting harm; all alternatives have been considered and rejected on documented grounds

3. Unavoidability confirmed - harm is genuinely unavoidable through the two-stage avoidability test, not merely inconvenient to avoid

4. Decisive asymmetry - expected casualties are substantially lower - by a very large margin -than casualties prevented; harm prevented must vastly exceed harm caused

5. Irreversibility - the threatened harm cannot be mitigated by subsequent action

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

7. Highest-level authorisation - authorised by pre-designated authority at highest practicable level

8. Contemporaneous documentation and automatic review - factual basis recorded at time of decision; independent post-strike review automatically triggered

The override may NOT be invoked for: high-value targets (unless actively directing imminent attack), strategic degradation, routine battlefield advantage, political or morale considerations, speculative harm, or targets of opportunity. It applies rarely, if ever.

H. Two-Tier Accountability

Tier 1 - Criminal Liability: For grave violations where civilian presence was known and avoidable, intelligence was misrepresented, reckless disregard amounting to wilful blindness, emergency override was abused, documentation was falsified, or command responsibility failures occurred.

Tier 2 - Administrative Liability: Automatic review for procedural failures (failure to verify, take precautions, escalate, or document), with consequences including retraining, suspension, operational restrictions, or separation.

Accountability operates in the first instance through national military justice systems. FHI does not displace domestic jurisdiction but provides standards against which national proceedings can be measured.

I. Remedies for Victims

Civilians harmed in violation of FHI are entitled to acknowledgment, compensation, medical assistance, access to information and truth recovery, and assurances of non-repetition.

4. Academic Endorsement

The framework has received substantive engagement from leading scholars. Oxford Professor Henry Shue, whose work with David Wippman on protective proportionality is foundational in this field, concluded that FHI's insistence on assessing avoidability before proportionality is 'radically better' than prevailing approaches that 'plunge right into proportionality weighing.' Shue acknowledged that his own prior work 'raises the right problem but does not offer much of a solution’ - and that FHI does.

Cornell Professor Matthew Evangelista has described FHI as 'a kind of intermediate mechanism between the systemic-level enabling of atrocity and the rules of engagement.'

5. Structure of the Declaration

The Declaration consists of 19 articles in 8 Parts, plus an Interpretive Guidance Annex:

Part I: General Provisions (Articles 1–2)

Part II: Substantive Obligations (Articles 3–7)

Part III: Emergency Override (Article 8)

Part IV: Procedural Obligations (Articles 9–11)

Part V: Remedies for Victims (Article 12)

Part VI: Relationship with International Law (Article 13)

Part VII: Implementation (Articles 14–16)

Part VIII: Norm Development and Final Provisions (Articles 17–19)

6. Compatibility with International Humanitarian Law

It is axiomatic that FHI must not and does not make new law. FHI makes explicit what current law implicitly demands; it does not amend IHL; it clarifies it.

• It is fully consistent with the Geneva Conventions, AP I Articles 51 and 57, the Martens Clause, and relevant jurisprudence.

• It aligns IHL's existing precautionary duties with modern technological reality.

• It operationalises 'constant care' and 'feasible precautions' by giving them real legal consequence.

• It reflects the protection-favouring principle recognised by courts and ICRC Commentary.

• It modernises, but does not contradict, proportionality.

FHI does not require amendment to the ICRC's formulation of customary rules. Rule 15's requirement to take 'all feasible precautions' to 'avoid, and in any event to minimize' civilian harm already contains the interpretive space FHI occupies. What FHI provides is operational specificity: it clarifies that where avoidance is feasible, minimisation is insufficient.

The 2023 US DoD Law of War Manual revision demonstrates that FHI's core requirements—including presumption of civilian status and verification obligations—are operationally achievable. The Manual's progressive language coexists, however, with permissive targeting practices; what FHI provides is the binding force that doctrine alone has lacked.

FHI engages with leading IHL scholarship, including Ohlin's two-track system (which it completes by adding the avoidability inquiry), Crawford's critique of systemic collateral damage (which it operationalises), and Shue & Wippman's protective proportionality concept (which it extends).

7. Why Governments Should Support FHI

Moral and humanitarian leadership: Demonstrates responsible behaviour and alignment with public conscience.

Strategic advantages: Reduces civilian harm, which fuels instability and delegitimises operations.

Technological realism: Modern capabilities already allow avoidability determinations.

Legal coherence: Clarifies ambiguities and reduces litigation risk by defining feasible alternatives and avoidability.

Political credibility: Provides a clear, defendable standard in parliamentary and congressional oversight.

Operational clarity: Gives commanders a predictable, evidence-based decision framework instead of vague balancing tests.

8. Why NGOs and UN Actors Should Support FHI

FHI:

• Provides a measurable, enforceable civilian-protection standard

• Elevates children's protection

• Aligns with the Political Declaration on Explosive Weapons in Populated Areas (EWIPA), the Safe Schools Declaration, and UN protection-of-civilians efforts

• Empowers investigative organisations

• Gives courts a concrete interpretive instrument

• Includes victim remedies often absent from IHL discourse

• Strengthens accountability pathways

9. Path to Adoption

Public Launch: Publish Declaration, Explanatory Report, Concept Paper, Policy Brief, and academic article. Seek NGO support.

Diplomatic Engagement: Present to Geneva missions. Engage states with strong protection-of-civilians records.

Integration into Doctrine: Embed the avoidability gate and capacity-indexed duties in manuals and ROE.

Judicial Uptake: Encourage courts to reference FHI's structured tests.

Long-Term Goal: Transition to an FHI Convention or Additional Protocol.

Conclusion

The Declaration on Foreseeable Harm to Innocents modernises IHL for an age in which civilian harm is increasingly foreseeable and increasingly avoidable. By introducing the avoidability gate and clarifying capacity-indexed duties, FHI restores moral and legal coherence to civilian-protection practice.

In modern warfare, preventable civilian deaths are not unavoidable.

FHI helps ensure they are avoided.