From Commitment to Practice

This article argues that endorsing the EWIPA Political Declaration is insufficient on its own to prevent civilian harm from explosive weapons in populated areas. Drawing on case studies involving endorsing states, it shows how even precision-guided munitions continue to cause severe civilian casualties due to inherent blast and fragmentation effects and contextual factors. The authors emphasise the need for evidence-driven implementation, including restraint and avoidance, weapon-specific restrictions, systematic civilian harm tracking, and learning processes that translate data into concrete operational and policy change.

Introduction

Since the Political Declaration on Strengthening the Protection of Civilians from the Humanitarian Consequences Arising from the Use of Explosive Weapons in Populated Areas was opened for adoption in 2022, 90 states have endorsed it. In doing so, they committed to strengthening the protection of civilians by restricting the use of explosive weapons in populated areas (EWIPA) and by addressing the humanitarian consequences when such weapons are used.

In the years since the Declaration’s adoption, however, the use of explosive weapons has continued at unprecedented levels. While much of this increase has been driven by conflicts involving states that have not endorsed the Declaration, civilian harm has also continued to occur in operations conducted by endorsing states. This underscores a central premise of the Declaration. Civilian harm from the use of explosive weapons by state actors is not confined only to situations of disregard for legal norms but can also arise even where states emphasise compliance with international humanitarian law, conduct attacks on legitimate military targets, and use advanced targeting and weapon systems.

This reflects a fundamental characteristic of explosive weapons. By their nature, many generate blast and fragmentation effects that can extend beyond the immediate target area. In populated and urban environments, these effects are further shaped by the built environment, where structures may sometimes shield civilians but can also amplify or redirect explosive force, trigger secondary explosions, and cause cascading damage to civilian infrastructure. As a result, the use of explosive weapons in populated areas carries inherent and foreseeable risks to civilians.

It is in this context that the Declaration’s added value becomes clear. International humanitarian law establishes essential legal obligations, but it does not prescribe how states should account for the effects of explosive weapons in populated areas, how urban warfare should be conducted, or what concrete measures should be put in place to prevent or mitigate civilian harm. These issues are addressed through national policies, doctrine, targeting procedures, training, and systems for tracking, analysing, and learning from harm. 

Yet implementation discussions so far suggest that some states continue to rely heavily on assertions of legal compliance or the sufficiency of existing policies, rather than engaging in detailed examination of how EWIPA-specific risks are managed in practice.

This article examines what effective implementation of the Declaration requires in light of these challenges. It first presents three incidents in which the use of explosive weapons in populated areas by endorsing states to the EWIPA declaration resulted in significant civilian harm. The purpose is not to condemn individual states reportedly responsible for this harm, but instead to illustrate recurring harm factors and to demonstrate why endorsement and procedural steps alone are insufficient to prevent civilian harm.

The article then explores how evidence from these cases can inform more effective implementation using the military implementation toolkit¹ developed by Article 36 and Airwars as an analytical framework. It examines how national policies and practices can be strengthened to prevent harm – through restraint and avoidance, weapon-specific limitations, improved civilian harm tracking, and learning processes that translate evidence into operational and policy change. In doing so, it aims to support constructive, evidence-informed dialogue on closing the gap between political commitment and practice, and on ensuring that the Declaration remains grounded in the realities it was designed to address.

Incidents of civilian harm from the use of explosive weapons – case studies²

Airstrikes on housing in Sa’ada, Yemen, by US armed forces, March 2025

On the night of 15 March 2025, an alleged strike by US armed forces during Operation Rough Rider reportedly killed at least 10 civilians, including a woman and nine children in a neighbourhood of Saada City, in northwest Yemen.³ Two residential houses were destroyed in the attack. 

Based on debris photographed at the site, the attack was conducted with Tomahawk Land Attack Missiles (TLAMs).⁴ These are American-designed long-range cruise missiles often launched from naval vessels. They employ four different guidance systems to reach predesignated target coordinates: GPS, inertial navigation, Terrain Contour Matching and Digital Scene Matching.⁵ Yet despite this sophisticated array of precision guidance systems, the use of an explosive weapon with such a large explosive load led to severe civilian harm when it was used in a neighbourhood with civilian residences. 

Endorsers of the EWIPA Political Declaration should look to restrict and refrain the use of heavy munitions such as the Tomahawk in populated areas. The high-explosive-power and blast-fragmentation warhead used in the Tomahawk make it highly prone to producing wide-area effects. The presence of 120 kg of high explosives in the warhead can also lead to significant overpressure effects from the explosives, as well as fragmentation from the bomb casing and materials in the surrounding environment. The use of such a weapon in a populated area can therefore negate many of the alleged benefits offered by precision as the impacted area often extends far beyond the original target. 

Use of drones by Turkish and Somali armed forces, March 2024

As part of Türkiye’s ongoing military operations in support of Somali armed forces, Bayraktar TB2 UCAVs have been employed to conduct strikes against suspected militants. On 18 March 2024, a farm in the Lower Shabelle region was struck during a Somali military operation in which Turkish drones were used. It remains unclear which states’ armed forces were operating the drones at the time of the airstrike.⁶ At least 23 civilians were reportedly killed in the attacks, with others injured. Somali forces later claimed they were targeting Al-Shabaab fighters in the area.

Debris analysed by Amnesty International found that Turkish-designed MAM-L air-delivered bombs were also employed. The MAM-L, which can be launched by Bayraktar TB2 drones, has been extensively documented by the Open Source Munitions Portal (OSMP) during its use across the region.⁷ The munition itself has a relatively small weight of 22 kg and can use a number of warhead variants, including blast-fragmentation, anti-tank, and thermobaric. Guidance primarily relies on drone operators painting a target using onboard laser designators, while GPS and inertial navigation provide an additional layer of control. 

Even with these systems, this incident exemplifies civilian harm from munitions equipped with smaller warheads and precision guidance, possibly based on faulty intelligence and an over-reliance on precision as an inherent mitigator. While explosive weapons with large warheads are often associated with wide-area effects, smaller munitions can also cause serious civilian harm when used in populated areas. Blast and fragmentation effects are relative to the target and surrounding environment, and can extend beyond the point of impact, affecting civilians and civilian objects nearby.

Airstrike in Iraq by Dutch armed forces, June 2015

During the night of the 2nd June 2015, an airstrike by Dutch armed forces targeting the Islamic State in Hawija, Iraq, killed at least 70 civilians alongside a large number of militants.⁸ The airstrike was targeting a car-bomb factory; ss a result of the initial strike, approximately 18,000 kg of TNT ignited, flattening many buildings in the surrounding area that were hosting residents and displaced Iraqi civilians.⁹ After a major investigation by Dutch media forced the Dutch Ministry of Defence to accept responsibility four years later in 2019, it was revealed that the Netherlands made the decision to conduct the strike as they had small and precise munitions. Five GBU-39 Small Diameter Bombs (SDBs) and one 500-pound bomb were used in the strike.¹⁰

The GBU-39 SDB is an American-designed GPS and inertially guided air-delivered bomb in the 250-pound weight class.¹¹ Its warhead is relatively small, containing 37 pounds (16.8 kg) of high explosive composition. The United States Air Force argues that the GBU-39 bomb’s small size “inherently reduces the probability of collateral damage”.¹² One variant, called the GBU/39A/B Focused Lethality Munition (FLM), was even developed to use a ‘focused-blast’ explosive warhead to reduce fragmentation spread during strikes¹³ and has been referred to as an “ultra-low collateral damage” munition.¹⁴

The presence of small and precise munitions failed to alleviate the risk to civilians in this instance due to a number of factors, some of which are often under-appreciated in planning. Notably, regardless of whether a munition is precise or not, the detonation of the contents of an IED factory negates precise targeting.

As a result of their involvement in this airstrike, the Dutch MoD undertook significant reform of its civilian harm mitigation and response mechanism. In the last five years, it has drastically improved its transparency mechanisms, its public reporting system, and its investigatory capacity.

Why the case studies matter for implementation

The case studies presented above are illustrative of patterns of harm from the use of explosive weapons in populated areas. Yet, they are far from the exceptions. They illustrate clearly that while endorsement of the EWIPA Political Declaration signals an important recognition of the risks posed by explosive weapons in populated areas, endorsement alone does not automatically lead to meaningful change.

The mechanisms through which harm arises – use of explosive weapons in populated environments, the projection of blast and fragmentation across an area, damage to civilian infrastructure – are precisely those the Declaration was designed to address. Its effectiveness therefore depends on whether implementation moves beyond formal endorsement and reaches the level where it influences operational decisions, which include promoting restraint and avoidance, prioritising protection of civilians, strengthening learning from harm, and ultimately preventing civilian harm.

States should proactively engage with this process, including by developing deliberate strategies to operationalise the Declaration’s specific commitments and understanding and managing the technical factors that influence harm. These factors include weapons effects, the projection of blast and fragmentation which is especially complex in urban environments, the risk of secondary explosions, and navigating damage to civilian infrastructure.

Procedural foundations

Surveys of endorsing states of the Declaration and statements delivered at the two implementation conferences that have taken place since 2022 clearly illustrate that many states have taken important first steps towards strengthening policy and practice, illustrating that implementation is increasingly understood as a national policy development process aimed at guiding operational practice.¹⁵ However, the depth and nature of implementation of the Declaration across national contexts varies. 

Examples of national policy measures relevant to EWIPA Declaration implementation¹⁶

Taken together, these examples illustrate how implementation can move beyond procedural changes to the development of new policies that emphasise restraint, provide clearer guidance, and put in place restrictions and measures to promote civilian protection in practice.

These steps are important for embedding the Declaration domestically and creating the structures for overseeing and sustaining engagement. Bureaucratic and procedural activity, however, will not, on its own, change operational outcomes or immediately reduce civilian harm. Procedural changes need to lead to more substantive changes. Recent implementation discussions suggest a tendency among some states to frame implementation in broad, high-level terms rather than engaging with operational and technical detail. This sits uneasily alongside the concrete and recurring patterns of harm documented in the cases above, including a continued reliance on pre-existing policies or legal frameworks as sufficient risks, without confronting evidence of continuing harm, thereby insulating the Declaration from the lived realities of civilians affected by explosive weapons. As illustrated by the Hawija case, meaningful reductions in civilian harm emerged not from procedural alignment alone but from subsequent scrutiny, learning, and policy reform following the recognition of harm.

Maintaining an evidence-driven implementation focus

A defining feature of effective implementation is the extent to which data on civilian harm and on weapons effects is used to shape policy and operational decision-making. Evidence-driven implementation requires that data on harm, including civilian casualties and damage to infrastructure - gathered internally and through external sources - actively informs decisions about restraint, such as policies to limit certain weapons or use in certain contexts, or authorisation levels, and learning processes. As the examples above demonstrate, such as in the case of the Netherlands, where evidence of harm has been acknowledged and examined, implementation is more likely to translate into substantive policy change grounded in the realities of civilians living through armed conflict. This requires openness, scrutiny, and a willingness to engage with difficult questions, rather than reliance on assertions of compliance of international humanitarian law or the sufficiency of existing policies, which is a tendency among some states.

As the case studies illustrate, civilian harm has at times been identified primarily through independent investigations by media outlets, independent monitors, and civil society, rather than through official state-led processes. This underscores the importance of transparent harm identification and engagement with external evidence as a foundation for learning and prevention.

Ultimately, the Declaration will be judged by whether it leads to decisions that prevent civilian harm, not by assurances that current approaches are adequate. Achieving this requires sustained, technical, and practical engagement – ideally through open dialogue among states, armed forces, international organisations, and civil society – focused on how harm occurs and how it can be prevented in practice.

Recommendations and toolkit

From evidence to action - implementing the Political Declaration in practice

The case studies and data examined in this article underscore why the EWIPA Political Declaration places such emphasis on learning from civilian harm and translating that learning into concrete changes in policy and practice. Endorsement of the Declaration reflects recognition of the humanitarian risks posed by explosive weapons in populated areas and a commitment to address them. Fulfilling that commitment requires implementation that is deliberate, evidence-driven, and oriented toward prevention.

The Declaration’s effectiveness will therefore depend on whether states use available evidence – including open-source information – to change how military decisions are made and how risks to civilians are addressed in operational settings. The military implementation toolkit developed by Article 36 and Airwars provides a practical framework to support this process. Drawing on the patterns of harm and operational challenges illustrated in the case studies above, several priority areas for action emerge.

Systematic tracking and analysis of civilian harm

Civilian harm tracking is the foundation of all mitigation efforts, as understanding the real impact of explosive weapon use is essential to identifying causes and preventing recurrence. However, in spite of growing recognition of the value and importance of armed forces tracking civilian harm resulting from their military operations, most armed forces still do not routinely track or investigate such harm.

It is vital that states establish mechanisms to record, investigate, and analyse civilian harm in real time, and ensure that findings are integrated into operational decision-making and, particularly, lessons-learned processes to drive sustained change. The lessons learned from civilian harm tracking should induce prompt and agile adaptations from militaries, as well as improvements to policy, doctrine, and training for future operations. Tracking mechanisms should actively feed into other civilian harm mitigation mechanisms, including investigations and response.

Militaries and states must proactively engage in dialogue with third parties (such as civil society organisations) and establish mechanisms for trust-based collaborative exercises in order to enable such organisations to support efforts on transparency, strengthening data quality, and supporting accountability.

Restricting and refraining from use

The above case studies underscore that even precision-guided munitions can cause severe civilian harm when used in populated areas, reinforcing the need for implementation approaches that prioritise restraint and, where appropriate, avoidance. Effective implementation must therefore include decisions not to use explosive weapons where civilian harm is foreseeable, supported by policies that explicitly address whether, when, and under what conditions particular weapons may be employed in populated environments. This, in turn, requires a strengthened understanding of the effects of different explosive weapons, including blast and fragmentation characteristics, wide-area effects, and the risks of secondary explosions in urban settings. Such analysis should inform explicit limitations on certain weapons, whether through avoidance policies, conditional use restrictions, or enhanced oversight and mitigation requirements. Where the use of explosive weapons in populated areas is nevertheless contemplated, authorisation frameworks should reflect the heightened risk to civilians, for example, through higher levels of approval, additional scrutiny, and weapon- and context-specific mitigation measures.

Conclusion

The continued and widespread civilian harm resulting from the use of explosive weapons in populated areas confirms the ongoing relevance of the EWIPA Political Declaration. While endorsement is an important starting point, the most meaningful change comes from the effective operationalisation of its commitments at the national level. This requires a strong focus on prevention, including building a culture of protection of civilians and moving away from the routine use of explosive weapons in cities, towns, and other populated areas. Ultimately, the credibility and impact of both the Declaration and its endorsing states depend on remaining closely connected to, and grounded in, the realities of harm experienced by civilians.