Rising Al-Shabaab Violence in Somalia’s Lower Juba Following Government Offensive

Somalia continued to experience intense battles between government forces and the armed non-state actor al-Shabaab in 2024, although ACLED data showed a decrease in civilian fatalities and injuries perpetrated by al-Shabaab compared to 2023.¹⁸ These clashes had a devastating impact on civilian lives and property, displacing thousands in affected areas. In one deadly assault against a beach hotel on August 2, al-Shabaab killed at least 37 civilians and injured more than 200 others.¹⁹

Compared to the rest of the country, the Lower Juba region experienced a sharp rise in violence in 2024, with increased attacks against civilians.²⁰ This escalation followed operations launched in early 2024 by Somali National Army (SNA) and Darawish Jubaland forces in Kismayo and its surrounding areas. The military campaign aimed to reclaim al-Shabaab-held territory and stabilize Kismayo and the corridor between Kismayo and Afmadow, a critical supply and movement corridor. Another key military objective was to reopen major supply lines linking the three main Jubaland-controlled towns of Kismayo, Afmadow, and Dhobley.

The joint SNA and Darawish operations from March to July 2024, succeeded in pushing al-Shabaab out of several strategic towns and villages around Kismayo. However, in retaliation, al-Shabaab launched a coordinated, multi-pronged offensive in July 2024, targeting SNA and Darawish bases in the recovered areas of Bulo-Haji and the towns of Harbole, Miido, and Bar Sanguni.²¹ These military operations and the resulting al-Shabaab counterattacks inflicted heavy costs on civilians. Communities suffered significant loss of life, destruction of property, restricted humanitarian access, and mass displacement.²² Al-Shabaab also reportedly increased violence and targeting of civilians, accusing them of collaborating with government security forces.²³ This civilian harm fits a pattern that is all too common in war—militaries retake but cannot protect or hold recovered territory, exposing civilians to retaliation from non-state armed groups and leaving a wake of destruction.

Civilians fleeing frontline hostilities were also displaced by the violence. According to UNHCR’s Protection and Solutions Monitoring Network (PSMN), approximately 570 households, comprising 3,420 individuals, were displaced to Afmadow and Kismayo.²⁴ The Jubaland Commission for Refugees and IDPs (JUCRI) reported even higher figures, estimating 828 households (4,968 individuals) were displaced.²⁵ Civilians displaced by the military operations either joined existing IDP camps or established new settlements, worsening an already dire humanitarian situation marked by high internal displacement and dwindling humanitarian aid.

Counting the Blasts: The Civilian Burden of Explosive Weapons in an Unrelenting Year

This article was authored by Iain Overton and Niamh Gillen from Action on Armed Violence (AOAV)

The use of explosive weapons in conflict has continued to rise, with civilians once again bearing the overwhelming burden. In 2024, Action on Armed Violence (AOAV) recorded 67,026 casualties from explosive weapons across 9,553 incidents worldwide. Of those harmed, 59,524 were civilians. This figure marks the highest annual total of civilian casualties in AOAV’s records since systematic monitoring began in 2010.

It must be stressed, however, that these figures are drawn exclusively from English-language media reports. They represent only a fraction of the true toll. Many regions suffer from limited press access, censorship, or reporting fatigue that leads to unreported or undercounted on incidents. This is especially true for conflicts where international media presence is weak or restricted and in areas experiencing sustained bombardment, such as Gaza. The data should therefore be seen, not as comprehensive casualty totals. Still, the data reveal clear and consistent patterns of harm. AOAV’s methodology does not aim to capture every death or injury, but to highlight the recurrent and preventable civilian impact of explosive violence.

The horrors of such patterns cannot be denied. In populated areas, 95 percent of those killed or injured by explosive weapons in 2024 were civilians. This reflects a trend AOAV has observed for over a decade: over this longer time period, when these weapons are used in towns, cities, or encampments, nine out of ten casualties are non-combatants.

Nowhere was this harm more evident than in Gaza. In 2024, AOAV recorded 23,432 civilian casualties in the enclave, including 14,145 fatalities. This figure represents 39 percent of all civilian harm from explosive weapons globally last year. Again, these numbers are drawn from English-language media and are likely to represent a significant undercount of the real toll. AOAV’s research suggests that as much as two-thirds of deaths in Gaza are not captured by such reporting. Nonetheless, the data point to a sustained and disproportionate pattern of violence against civilians. Almost all recorded incidents in Gaza took place in populated areas, and 99.9 percent of civilian casualties were attributed to Israeli state forces.

Beyond Gaza, Lebanon experienced a dramatic escalation. Civilian casualties there surged by more than 6,000 percent, from 141 in 2023 to 9,100 in 2024. A large part of this increase was driven by Israeli use of explosive devices disguised as pagers and radios in densely populated areas of Beirut and southern Lebanon. These attacks represent one of the deadliest uses of improvised explosive devices (IEDs) by a state actor in AOAV’s dataset.

Ukraine continued to face high levels of civilian harm. AOAV recorded 11,765 civilian casualties there, making it the second most affected country after Gaza. Nearly all of these casualties occurred in populated areas. Although air strikes rose significantly, the majority of civilian harm was still caused by ground-launched weapons, which tend to be indiscriminate in their effect. In Sudan, where the civil war deepened, AOAV recorded 4,478 civilian casualties. This marks the highest total in Sudan since AOAV began recording, with many attacks targeting residential areas, markets, and encampments.

In Myanmar, explosive violence killed or injured 3,379 civilians, a 56 percent increase from 2023. Most of this harm was caused by the military junta’s aerial bombardments. These attacks tended to hit urban areas and often targeted infrastructure vital to civilian life. In Iran, a single suicide bombing at a public gathering caused hundreds of casualties, contributing to a recorded 8,320 percent rise in civilian harm from explosive weapons in that country.

AOAV’s data reveal, not just the scale, but the nature of explosive violence in 2024. Of the 9,553 recorded incidents, 86 percent occurred in populated areas, while 97 percent of the casualties from these incidents were in populated areas. Residential zones, schools, camps for the displaced, and urban marketplaces were frequently targeted. In total, 8,204 such incidents were recorded in populated settings, causing the deaths or injuries of 57,506 civilians.

In 2024, explosive violence was dominated by state actors: states were the reported perpetrators of 7,999 incidents, which resulted in 56,465 total casualties across 28 countries. Of these casualties, 93 percent (52,259) were reported as civilians. Israel was the predominant state perpetrator of civilian harm, accounting for 33,372 civilian casualties, which is 55 percent of all the civilian casualties recorded globally in 2024. These civilian casualties were recorded across seven states and territories, but 70 percent (23,425) of the civilian casualties caused by Israel took place in Gaza. 

In the first five months of 2025 (January to May), AOAV has recorded 3,371 incidents of explosive violence, resulting in 18,999 civilian casualties (7,869 killed). Gaza remains the worst impacted, with 31 percent (5,937) civilian casualties recorded, followed by Ukraine where 28 percent (5,363) of civilian casualties were recorded. This is a similar level of harm in comparison to the same time period in 2024, but with levels of civilian harm showing little sign of reducing in the near future, it can be expected that these figures could surpass the levels seen in 2024. 

Our evidence points to an overwhelming truth. Civilians are consistently and disproportionately affected when explosive weapons are used in populated areas. Despite the adoption of the 2022 Political Declaration on Explosive Weapons in Populated Areas, little has changed on the ground. State and non-state actors continue to use wide-area effect weapons in locations where civilians live, work, and seek refuge.

This report is not a complete ledger of loss. But it does serve as a warning as to what happens when war is fought in streets, shelters, and homes. And it calls on those with power and responsibility to change course from such behavior.

A Civilian First Line of Response: The Growing Importance of Community-Based Protection

When conflicts erupt, civilians are often the first line of response to protection threats. Communities, families, and individuals engage in their own forms of self-protection and community protection, particularly in situations where there is limited state capacity or presence. Such community-based protection (CBP) measures are regularly neglected by external protection actors and donors but have become increasingly recognized in recent years. There is a growing realization that integrating CBP into programming is often the most effective and sustainable response to crises, as it taps into the capacities, knowledge, and agency that already exist at a local level.⁵⁷ The humanitarian funding crisis that began in 2024 has added new impetus: as resources dwindle and access to communities becomes more difficult, CBP can offer a more cost-effective and locally-grounded way of responding to protection needs.⁵⁸

CBP approaches can include a wide array of activities. One key function is to empower communities to lead on engagements with armed actors before, during, and after a conflict. Such interactions often happen out of necessity at the community level before any external actors have time to deploy to support more formalized civil-military engagement, mediation, or negotiations. Engagement between civilians and combatants is complex and not always clearcut, particularly because the lines between armed actors and communities can be blurred. Individuals might play roles as influential civilian figures when out of uniform but also operate as part of an armed force. Overlapping social bonds can create an organic environment for negotiations.⁵⁹ External peace actors can play a crucial role in equipping communities with tools to support such interactions, including through mediation training or by providing platforms for dialogue.

A recent example comes from Yemen, where millions have been displaced and more than 12,400 civilians killed in conflict since 2015. Although high-intensity violence has subsided in some parts of the country, it has continued in areas such as al-Dhale’e, Marib, Taiz, and Lahj. Sporadic clashes and military operations have caused civilian deaths and injuries, and limited movements in search of healthcare or livelihood. To mitigate these harms, CIVIC has worked with local actors to establish Community Protection Groups (CPGs) since 2018 that support communities to engage directly with armed actors on reducing harms. The CPGs are rooted in Yemeni community traditions, and composed of local civil society actors and other leaders, all with a direct tie to the areas in which they work.

These CPGs have taken on a range of functions, including: collecting information on civilian harms at a local level; developing proposed mitigation measures; public awareness raising through media outreach and other events; and strengthening local knowledge of the importance of POC, with a focus on the role of women. In addition, CPGs have brought communities and armed actors together to explore how to collaborate on reducing civilian harms. Such meetings have also provided a forum for raising specific abuses, including arbitrary detentions or harassment at checkpoints. There have been some notable results, including in local dispute resolutions, and in reopening of roads previously closed due to conflict. Such road closures had forced community members to take dangerous alternative routes, while limiting trade and humanitarian access, driving up commodity prices.

Likewise, in 2024, CIVIC-supported CPGs in Nigeria changed military policy and practice through civil-military dialogue in certain conflict-affected areas of the Northeast, including by reversing a harmful decision to ban the use of bicycles on farms, widening the protected town perimeters to create more farming space to accommodate growing town populations, directing military patrols to high-risk areas, and removing soldiers accused of committing sexual exploitation and abuse for follow-up disciplinary action.

Beyond engagement with armed actors, CBP spans a range of other activities, such as efforts to strengthen community preparedness for violence and address social cohesion concerns that drive escalating violence or ignite tensions within communities. In Ukraine, for example, CIVIC worked with CPGs in frontline communities in 2024 to draft Emergency Road Maps outlining activities and roles for responding to escalating security crises in ways that reduce protection threats. Capacity-building on conflict-sensitive communications for CPGs has helped resolve disputes between civilians and demobilized military personnel, or between different social groups (for example, older persons, IDPs, families with mobilized relatives, and parents of children with disabilities.) 

Additionally, community-level work is often key to monitoring violations and raising early warning concerns. For example, Comités de veille (“Monitoring committees”) in the Timbuktu, Djenné, and Mopti regions of Mali have facilitated dialogue and information-sharing on protection concerns with local authorities, helping  highlight risks faced by civilians and support more responsive security measures. In the DRC, protection monitors work in collaboration with humanitarian actors to identify and report threats, while local watch groups alert authorities to security incidents. The ongoing withdrawal of MONUSCO, however, threatens to undermine some of these systems, as the Mission supported Community Liaison Assistants—who played a key role in liaising between state authorities, armed groups, and villages on security risks—have exited some communities (See Spotlight: UN peacekeeping withdrawals).

Moreover, CBP approaches can provide protection for at-risk or marginalized groups. In South Sudan, Nonviolent Peaceforce has helped establish Women’s Protection Teams that created child-friendly spaces in their communities and played key roles in spreading awareness of waterborne diseases caused by regular flooding.⁶⁰ Indeed, such climate-focused measures are increasingly becoming a feature of CBP efforts. In the Philippines, peacebuilders have worked with forest rangers to better detect illegal logging and engage with illegal loggers through nonviolent communication to reduce tension. These efforts have also focused on more sustainable ways to manage natural resources, while feeding into early warning systems focused on disaster risk reduction.⁶¹

Self-protection initiatives can be an effective way to reduce harm, but they are not without risks, especially when they involve engagement with armed actors. In central Mali, for example, CBP interventions in 2024 allowed communities to negotiate agreements with NSAGs to ensure continued access to markets, livelihoods, and livestock grazing. Such engagement, however, risks falsely stigmatizing communities in the eyes of security forces who suspect them of supporting armed groups. This points to a key challenge in CBP programming: the need to develop risk mitigation measures and ensure that international actors prioritize the safety of local partners. In Ukraine, for example, national frontline responders often absorb the majority of the risks as they deliver aid or engage in protection efforts in areas where INGOs have deemed it too dangerous to deploy international staff. International actors have been criticized for not providing local “subcontractors” with adequate security or protection, effectively failing to meet their duty of care. Recent research by ACAPS and Nonviolent Peaceforce stressed the immediate, practical steps international actors should take, including providing PPE, covering the costs for fuel and vehicle repairs, providing insurance coverage, as well as providing training in safety and security, and relevant international law and standards.⁶²

Moving forward, it is crucial that donors, governments, and practitioners across the humanitarian-peace-development nexus build on recent progress to better integrate CBP approaches across interventions. As a coalition of NGOs led by ICVA recently stressed: in a time of drastically shrinking aid budgets, CBP is not just an ethical imperative but an operational necessity.⁶³ It is crucial that donors ensure that these efforts are prioritized through flexible, multi-year funding, to give efforts time to produce results in the medium- to longer-term. Crucially, this must be done as genuine partnerships where community-based actors have agency and decision-making power, and are not merely subcontractors. Inclusivity is also key, as CBP efforts should include all groups, including marginalized and at-risk members of the community, not just established leaders. Finally, it is important to stress that CBP does not mean leaving local actors to fend for themselves. International partners must establish complementary practices that work in tandem to reinforce, not replace, local efforts.

Managing the Protection Gap: Challenges from Abrupt Peace Operations Withdrawals

Analysis for this spotlight article is based on a more comprehensive policy brief published by CIVIC, the Norwegian Refugee Council (NRC), and the Global Protection Cluster (GPC) in 2024, titled Protection After UN Peacekeeping Mission Departures.⁷⁸

As noted, UN peace operations have faced decreased budgets and political support in recent years, leading to withdrawal of staff from several countries, even in contexts where conflict and violations against civilians have continued. Because modern peacekeeping missions are multi-dimensional with expansive mandates, their withdrawals have wide-ranging implications in insecure contexts. Most obviously, the removal of armed blue helmets from an active conflict context can create a major gap in the physical protection of civilians. This “security cliff” is particularly challenging because few other protection actors have the mandates or resources to replicate the range of activities missions undertake—from policing, to training security forces, to logistics. Mission withdrawals also mean that community-based protection models, including early warning systems—such as those developed by missions in the CAR, the DRC, Mali, and South Sudan—may be lost. In Mali, for example, MINUSMA had no time to transfer management of its early warning system amid its rapid withdrawal, while in the DRC, there are concerns about the capacity and funding of state actors to oversee the system MONUSCO is expected to hand over.

The departure of UN missions’ Special Representatives of the Secretary-General (SRSGs) reduces high-level political engagement on protection with national and regional senior leadership. In terms of humanitarian coordination, mission withdrawals can have significant consequences—including the withdrawal of mission POC advisors from the humanitarian cluster system, as well as diminished capacity to support the development of comprehensive national protection strategies. The implications for human rights monitoring can also be severe, as in the case of Mali, where the national government opposed a continued OHCHR presence after MINUSMA’s withdrawal. Even in cases where OHCHR maintains staff in areas where other mission components are no longer present, such as the DRC, they often face reduced resources and access, while risks of reprisals mean it is challenging for national actors to fill this gap. Peace operation closures also have implications for other actors in the protection field—particularly mine action and disarmament, demobilization, and reintegration (DDR) initiatives. Non-mission protection actors also often benefit from the security and logistics support of missions, and their closure can have significant implications for humanitarian access, especially if accompanied by drastic funding cuts to the UN Humanitarian Air Service (UNHAS).

At a national level, there are immediate, practical steps donors, protection actors, governments, and others can explore to mitigate the protection risks posed by abrupt withdrawals. These include, for example, strengthening dialogue with armed actors on protection, alongside earlier and ongoing engagement with national authorities on early warning systems during mission deployment. The UN could also consider expanding Resident Coordinator mandates to include engagement with high-level government officials on protection, while also mainstreaming and embedding protection analysis and other functions into UN Country Team frameworks.

More broadly, mission withdrawals could leave remaining protection actors facing a “financial cliff” as they become dependent on less predictable funding streams. To mitigate this, protection actors should consider advocating for more flexible funding arrangements after withdrawals, such as those employed in South Sudan through the Reconciliation, Stabilization, and Resilience Trust Fund (RSRTF). To ease the additional demands on humanitarians following mission departures, protection actors could engage UN entities and donors on options for deploying surge capacities of protection personnel in transition contexts. Some standing capacities already exist through the IASC and UNMAS, which could be leveraged or expanded. Finally, it is crucial to ensure that even in the context of short-term withdrawals, UN missions engage in a structured way with national protection actors on transition plans. In 2024, for example, MONUSCO’s engagement with non-UN actors was limited and began very late in the process of its withdrawal from South Kivu, leading to frustration among NGOs. This can be mitigated by the UN Security Council stressing integrated planning, data-sharing protocols, and early engagement with non-UN actors in mission withdrawal protocols.

From Policy to Practice: Military Efforts to Operationalize the Protection of Civilians

Improving civilian protection requires armed actors to operationalize legal, strategic, and ethical commitments into concrete practice. Many armed actors have taken steps to better embed POC efforts in military operations. Protection and civilian harm mitigation and response (CHMR) measures are wide-ranging and can include—but are not limited to—training of military personnel in IHL and related protection standards, establishing stronger monitoring and oversight mechanisms, improving accountability for violations, and increasing transparency in efforts to combat such harm. In 2024, analysts raised concerns that rapid global militarization and increased arms spending were rarely accompanied by an increased focus on integrating CHMR—although there were also notable positive developments.

The US, throughout 2024, continued to implement CHMR commitments established in the preceding years. Partly in response to a US airstrike in the Afghan capital of Kabul that killed 10 civilians in 2021, the Department of Defense (DoD) released a Civilian Harm Mitigation and Response Action Plan (CHMR-AP) in August 2022.⁸¹ The plan reflected an unprecedented commitment to overhauling US policies for preventing, mitigating, and responding to civilian harm resulting from US military operations. In the two years since its release, an assessment by CIVIC found that the DoD had made progress in putting in place staff and structures aimed at protecting civilians, but that there was still work to be done in implementing the CHMR-AP in practice.⁸²

By 2024, the DoD had hired almost all CHMR-related roles established under the plan and had set up various institutions across the Department to lead on CHMR work, including a new Civilian Protection Center of Excellence (CP CoE). At the combatant command (CCMD) level, new teams were set up to improve analysis of the civilian environment with a view to mitigating harm, as well as cells to oversee civilian harm assessments. The DoD has also initiated the development of new policies and procedures for responding to civilian harm, including acknowledgement, condolence or ex gratia payments, and in-kind and community-level responses. Their impact in practice, however, remains to be seen—despite a $3 million fund for ex-gratia payments, very few payments have been made over the last few years.⁸³

There were notable developments on CHMR in two other NATO member states in 2024. The Netherlands established a centralized civilian harm reporting portal, where civil society and affected civilians can submit reports of alleged harm by Dutch forces.⁸⁴ This portal builds on previous CHMR efforts by successive governments in the Netherlands, in large part motivated by Dutch air strikes in Iraq in 2019 that killed 89 civilians. Dutch policy, for example, now requires that the government proactively informs Parliament prior to military deployments about the anticipated risks to civilians and mitigation measures.⁸⁵ The Civilian Protection Monitor—an initiative launched in April 2024 by PAX and Airwars to monitor state policy and practice on protection of civilians—has noted, however, that these efforts mostly apply to Dutch soldiers deployed to international peacekeeping operations, but may not apply if the Netherlands participates in a collective defense operation under NATO’s Article 5.⁸⁶ The UK, conversely, regressed on CHMR and failed to institutionalize many policy commitments on the protection of civilians. Neither does the UK appear to consistently monitor or investigate reports of civilian harm to internationally recognized standards, while it has also eliminated mechanisms offering voluntary compensation payments to civilians harmed by military operations and introduced legislation to limit such payments.⁸⁷

The challenges of scaling CHMR policies from counterinsurgency to large-scale warfare came into focus last year, not least due to widespread atrocities in Ukraine since Russia’s full-scale invasion. This issue was also highlighted by the International Contact Group on Civilian Harm Mitigation and Response, an inter-state working group led by the US and the Netherlands, as most CHMR policies originated from counterinsurgency operations in countries such as Afghanistan and Syria.⁸⁸

The Yemeni military and security leadership from the Aden and Marib governorates advanced their country’s POC tools in March 2024 by signing a POC Code of Conduct Policy. In a November 2024 workshop supported by CIVIC, senior commanders from military and security institutions established an implementation plan that will include dissemination to all military and security sector leaders. It will also strengthen the oversight role of the military police to monitor implementation and establish a supervisory role in the security sectors for the same purpose. The aim is ultimately to involve communities themselves in the oversight and implementation of the plan. The initiative builds on previous engagement by peace actors, including CIVIC, focused on providing platforms for civil-military dialogues with civil society organizations and communities representatives, along with training of security forces in POC and CHMR.⁸⁹

The Iraqi Prime Minister approved the country’s first National Protection of Civilians Policy which explicitly commits the security and defense forces to respect a range of IHL and IHRL provisions as well as protection principles. These include: a prohibition on torture, excessive use of force, and the use of internationally banned weapons; a commitment to distinguish between combatants and civilians; and a duty to avoid excessive collateral damage. The policy also outlines objectives such as strengthening civilian harm mitigation strategies and mechanisms and striving to compensate civilians for harm.⁹⁰

In Nigeria, civilian deaths and injuries from military airstrikes have long been a concern. These were not due to deliberate targeting of civilians, but rather misidentification of targets by the military. Such mistakes result from both gaps in technical capacity in the military’s air-to-ground integration (AGI) and the considerable challenges of pursuing NSAG targets that are often closely integrated within local communities.⁹¹ In a positive development, the Nigerian military in 2024 took the rare step of prosecuting two officers over an airstrike that led to 80 civilian deaths, although at the end of the year, it had yet to follow up on promises to make public its own investigative report into the incident.⁹² While US support to the Nigerian military on strengthening CHMR protocols and AGI efforts have been positive, serious gaps remain, according to an assessment by CIVIC. These include a lack of institutional frameworks to apply best practices, scarce reliable intelligence, and the exclusion of military advisors from targeting processes.⁹³ In light of these challenges, other CHMR efforts are critical, including Nigerian military efforts to regularly train military personnel on POC and CHM, empowering military human rights and gender officers to monitor and address violations, and designing codes of conduct.

Moving forward, armed actors must better prioritize and resource CHMR efforts, especially given rapidly rising global arms spending. This should focus on establishing robust and transparent systems for mitigating military harm and ensuring these are incorporated across all military efforts. It is also crucial to ensure that existing CHMR policies are continually updated to reflect new developments in warfare, such as the risks posed by MDMH or the increasing use of explosive weapons in densely populated urban areas.⁹⁴