Hovering Threats: Managing the rising risks of armed drones in humanitarian operations

Christina Wille and Christa Callus from Insecurity Insight discuss some of the unique challenges posed by armed drones in humanitarian contexts, and suggest practical mitigation measures to counteract these challenges.

Armed drones are an escalating threat in humanitarian settings, with their use expected to grow. Once rare, they are now a defining feature in conflicts across Ukraine, the Occupied Palestinian Territories, Sudan, and Myanmar. Their appeal is clear: drones are low cost, easy to operate, widely available, and carry no pilot risk. But for humanitarians on the ground, they represent a deadly challenge – particularly when aid efforts take place near suspected insurgents.

Recent analysis by Insecurity Insight highlights how severe and fast-moving this threat has become. Between 2023 and 2024, the number of incidents in which aid or health operations were directly affected tripled, and the geographic spread expanded to at least 12 countries. Since 2016, at least 21 aid workers and 73 health workers have been reportedly killed by drone strikes – with over three-quarters of fatalities occurring in 2024, highlighting a relatively new and escalating concern.

The unique challenges posed by drones

Unlike traditional weapons, drones introduce a set of unique and fast-evolving challenges:

• Reduced warning times: The capability of drones to both gather intelligence and deliver precision or wide-area effect munitions within minutes severely reduces warning times. The effectiveness of early warning systems is also limited by the quiet, fast moving nature of some drones.
• Wide-area blast effects: Thermobaric and high-explosive warheads, deployed by some drones, produce wide blast effects that threaten not only the target but nearby infrastructure, civilians, and humanitarian operations.
• Misidentification risks: Aid vehicles, convoys, and even health facilities are increasingly being misidentified or framed as legitimate military targets.
• Autonomous targeting: Some drones can operate autonomously and self-select targets based on pre-programmed image recognition. There is usually no information on the image criteria used to train the target identification.
• Remotely piloted drones: Drone operators may have instructions to target operations of a particular nature, which may include aid operations. Other operators, while instructed to avoid targeting civilians or aid operations, may be poorly trained and have difficulties distinguishing between civilian and military targets.
• Drone technology changes fast: the risk environment is constantly changing – requiring constant adjustments.
• Deterrent effects: Even when drones do not strike, their constant presence in operational areas cause fear and programme suspensions – reducing access to critical aid.

Rethinking Security Risk Management

Adapting to the growing threat of drones means rethinking security risk measures in high-risk contexts. All mitigation measures must be tailored to the specific situation and these may change over time depending on the specific location. Insecurity Insight has published a practical guide that brings together current best practice, available in English and French.

Managing Movement and Visibility

Over half of aid worker fatalities from drones occurred during travel. Movement – particularly by vehicle – represents a significant vulnerability and requires careful planning.

• Limit non-essential travel, especially through high-risk areas such as those near militant activity, and areas surrounding contested sites (e.g., refugee camps, markets, or fuel depots).
• Avoid travelling during low visibility conditions such as dusk, fog, dust storms, or at night, when the risk of misidentification is higher.
• Choose vehicles with care:
  • Avoid those resembling military or insurgent vehicles, and be aware that their choice of car or visual identity may change over time.
  • Reassess the visibility of logos – a rooftop logo might offer protection in some settings but could draw attention in others.
  • Be cautious with armoured vehicles, which are generally less maneuverable, ineffective when directly targeted by first person view drones, and may look like a military vehicle.
  • External antennas and other visible communications equipment may mark vehicles as potential targets.

Strengthen organisational and staff awareness

Given the pace of technological change, drone threat monitoring should be integrated into regular context analysis and acceptance strategies. This includes:

• Keeping up with changes in local patterns of vehicle use, movement restrictions, and types of drones deployed.
• Regularly reassessing how an activity might look from the air. Many drone strikes occurred in the context of counter-terrorism or counter-insurgency campaigns. Humanitarian operators providing services to communities seen to be linked to opposing conflict actors are at risk of being labelled as collaborators themselves. Projects that attract large crowds, such as food or aid distributions, are also at risk of being misidentified and targeted.

A key part of organisational readiness is making sure all staff and partners know how to respond if a drone is present. This means incorporating drone-specific awareness into security briefings and trainings. General advice should include:

• Seek cover – indoors or at lower ground levels is safest.
• Stay still and lie flat – motion can attract the drone’s visual sensors.
• If no cover is available and the drone is close, run in a zig-zag pattern towards the nearest cover to make targeting more difficult.
• If you are in a group, disperse in multiple directions to reduce collective risk.
• Do not approach a landed drone – it may still be armed, with a kill zone of seven metres or more.
• Be aware that the drone operator may still be observing.

This can be handed out as a standalone document distinguishing between best practice inside (FR) and outside (FR).

Supporting programmes and local partners facing frontline exposure

Certain aid activities—particularly those that bring large groups of people together, such as food distributions, mobile health services, or emergency medical aid following an attack—may be at heightened risk. Security risk managers should proactively engage with programme teams and their local partners to ensure measures are in place to limit crowd sizes in open spaces where drone attacks are a concern.

Local and national partners often face heightened risks – not only while on duty, but also at home or during commutes, simply due to their proximity to conflict-affected communities. These risks are frequently overlooked in risk management plans.

Aid agencies should, wherever possible, offer partners the same level of security-related training and support as their own staff – helping to build shared capacity, promote good practices, and reduce risks linked to gaps in information or experience. Where possible, engage local partners in conversations about the security risk management support they need to safely carry out frontline work.

A collective challenge

Rapid advancements in drone technology make it difficult for individual aid agencies to keep pace alone. This is a collective challenge – one that requires shared analysis, collective approaches to mitigation strategies, joint advocacy, and sector-wide coordination.

Given the current political climate, it is more important than ever that the sector coordinates with legal and policy experts to strengthen protections under international law. The growing weaponisation of drones poses not just a security risk, but a direct and escalating threat to humanitarian operations. Autonomous weapons systems, in particular, raise serious questions around how principles of distinction and proportionality are upheld. Ultimately, even the most robust risk management strategies cannot safeguard humanitarian workers if warring parties do not respect International Humanitarian Law (IHL).

In parallel, the narratives that accompany drone use also matter. Disinformation campaigns have increasingly been weaponised to delegitimise humanitarian actors, portraying them as aligned with armed groups and undermining the value of aid.^[1] These false narratives not only put aid workers at greater risk of targeting – they also erode the neutrality and trust on which humanitarian access depends.

Finally, there is a need for effective accountability measures when drones are used to kill aid workers. There are clear risks for individual aid agencies in engaging on these topics – but silence comes at a cost. It is high time that the aid sector took a more collective stand in defending the humanitarian space by addressing technical issues around drones as well as the practices in accountability.

Additional resources

• Personal safety: Outside when a drone is spotted, available in EN, FR, and specifically for health care developed for the context in Myanmar also available in EN and Burmese
• Personal safety: Inside a building during a drone attack, available in EN, FR, and specifically for health care developed for the context in Myanmar in EN and Burmese
• Guidance on approaching a landed drone: available in EN, FR,
• Preparing travel, available in EN, and FR
• General guidance prepared for local partners across the Sahel in EN and FR
• General guidance prepared for health care in Myanmar in EN and Burmese.
• Security Incident Information Management (SIIM) can be used to collect and use information related to safety and security-related incidents to support and improve staff safety and security and improve access to affected populations.

The views and opinions expressed in this article are solely those of the author. They do not necessarily represent the views or position of GISF or the author’s employers. 

About the authors 

Christina Wille is a founding member and Director of Insecurity Insight. Her work focuses on improving data collection on violence and its consequences for humanitarian aid. A German and Swiss national, Christina holds an MPhil in International Relations / European Studies from the University of Cambridge.

Christa Callus is a senior researcher and data analyst with Insecurity Insight. She oversees data quality control for the organisation’s conflict incident database and manages Insecurity Insight’s work on the use of explosive weapons affecting humanitarian operations. She holds a Master in Humanitarian Action (NOHA) and is completing an MA in Human Rights Law at the University of Malta.

^[1] Insecurity Insight’s recent social media sentiment analysis noted that negative and inflammatory portrayal of aid, fuelled by the recent USAID Freeze, may heighten real-life security risks for organisations working in these settings.

Related: