Publication · April 19, 2026

Towards a Drone Non-Proliferation Framework (DNPF)

By Adithya Suresh · 26 min read

A structured proposal for a Drone Non-Proliferation Framework that addresses existing governance gaps.

Illustration of drones, international security, and governance
Drone policy and governance

The sky is no longer the limit. Drones have moved beyond military reconnaissance and conflict into civilian life, from consumer quadcopters to industrial aircraft. Their remote and autonomous operation is changing logistics, creative industries and warfare, making capabilities faster, cheaper and more accessible.

On modern battlefields, drone proliferation lowers the threshold for violence and increases civilian vulnerability. International frameworks designed before autonomy became widespread are no longer sufficient. A dedicated monitoring and verification regime is now a policy necessity.

A Brief Background & The Current Scenario

Cheap drones grew from electronics miniaturisation in the 2010s. Smartphone components—cameras, GPS, gyroscopes, accelerometers and powerful processors—combined with lighter, denser batteries and hobbyist innovation. Drone racing, open-source flight controllers and modified off-the-shelf components helped make capable systems widely available. Consumer prices range from about $25 for toy drones to $9,000 for prosumer models capable of speeds near 70 mph (112.65 km/h).

Military strategies have shifted from high-end surveillance UAVs toward swarms of small, inexpensive systems that can saturate air defences. A low-cost drone can threaten a high-value asset or overwhelm a costly defence system. Photography drones can also be modified for reconnaissance or attacks by attaching payloads. Thermal and infrared sensors, software and AI tools have expanded their usefulness at night.

The technology also reaches underwater. Uncrewed Underwater Vehicles (UUVs) can conduct surveillance, sabotage or kinetic strikes. Their low visibility and limited underwater situational awareness make them potential tools for deniable attacks on critical maritime infrastructure.

The Human Cost

Drones now reach civilian areas beyond front lines. Their persistent presence in cities creates a form of asymmetric psychological pressure. Low-cost loitering munitions can target power grids, water facilities and communications infrastructure; even when an attack is limited, its cumulative impact on daily life can be severe.

To civilians, a delivery drone and an armed FPV (First Person View) drone may be difficult to distinguish. Without regulation, everyday technology can become an instrument of harm, leaving urban populations exposed to constant uncertainty.

The psychological effects extend beyond physical strikes. The buzz of small motors can trigger hyper-vigilance and post-traumatic stress in communities exposed to repeated UAV activity. Operators also face a distinct ethical and psychological burden: they engage through high-definition screens while physically removed from combat. Research describes PTSD, anxiety, depression and moral injury among operators, alongside concerns that distance can weaken accountability and moral restraint. Drone proliferation therefore affects the human experience of conflict on both sides.

The Governance Gap

Existing legal frameworks remain tied to 20th-century weapon categories. This disconnect creates grey zones that state and non-state actors can exploit.

I. Under the Wassenaar Arrangement

The Wassenaar Arrangement promotes transparency and responsibility in transfers of conventional arms and dual-use goods. Its technical thresholds do not always fit low-cost or modified drones.

“7.A.3 ‘Inertial measurement equipment or systems’ … ‘Positional aiding references’ independently provide position and include ‘Satellite Navigation System’ and ‘Data-Based Referenced Navigation’ (DBRN).”

Controls on satellite navigation equipment address high-end GNSS, but electronic warfare has made GPS dependence a liability. DBRN allows systems to navigate by matching terrain rather than relying directly on satellites.

“7.E.4.B.5. Integration of digital flight control, navigation and propulsion control data, into a digital flight management system for ‘total control of flight’.”

Integrated flight management is relevant to civilian delivery and logistics as well as coordinated attacks. The same underlying capabilities can serve both purposes.

“8.A.2.h ‘Robots’ specially designed for underwater use … [with] the ability to exert a force of 250 N or more or a torque of 250 Nm or more.”

A force threshold aimed at heavy-duty robotic arms can miss low-torque underwater drones capable of damaging cables or pipelines with cutting tools.

“9.A.12. ‘Unmanned Aerial Vehicles’ … having a maximum endurance greater than or equal to 30 minutes.”

Many FPV attack drones operate for less than 30 minutes. Endurance thresholds can therefore leave effective tactical systems outside controls, while modular components make civilian-to-military conversion easier.

II. Under the Arms Trade Treaty (ATT) and the UN Charter

“The Treaty shall apply to all conventional arms within the following categories: battle tanks; armoured combat vehicles; large-calibre artillery systems; combat aircraft; attack helicopters; warships; missiles and missile launchers; and small arms and light weapons.”

The ATT does not explicitly define drones or UAVs. Exporters may classify them as civilian hardware or dual-use electronics, leaving strategic strike capability beyond the treaty's intended scrutiny.

“If the export is not prohibited under Article 6, each exporting state party … shall … assess the potential that the conventional arms or items would contribute to or undermine peace and security [or] be used to commit or facilitate serious violations of international humanitarian or human-rights law.”

Without a clear drone category, these risk assessments may not consistently cover unmanned systems. The treaty also lacks multilateral reporting and inspection mechanisms, relying substantially on exporters' own reporting.

“The States Parties shall make every effort to achieve consensus on each amendment … [and] as a last resort, an amendment shall be adopted by a three-quarters majority vote.”

Consensus is difficult in the current geopolitical climate, and the ATT's assumptions about high-value, state-controlled weapons do not map neatly onto cheap systems available to commercial buyers.

III. Additional Protocol IV on Blinding Laser Weapons

“It is prohibited to employ laser weapons specifically designed … to cause permanent blindness to unenhanced vision.”

Laser systems support target acquisition in autonomous kill cycles. Protocol IV should be considered alongside incidental blinding effects and laser systems used in targeting, while protecting civilians from harmful collateral exposure.

“Blinding as an incidental or collateral effect of the legitimate military employment of laser systems … is not covered by the prohibition of the protocol.”

Any DNPF should preserve these protections as automation expands and ensure that civilian safety is addressed in the design and use of laser-assisted systems.

IV. The UN Register of Conventional Arms (UNROCA)

UNROCA's combat-aircraft categories focus on manned aircraft and unmanned systems designed or modified to use missiles, rockets, bombs or guns. The definition should address attacks on civilian and critical infrastructure, non-state actors and the dual-use nature of unmanned platforms. Underwater drones and UUVs also remain outside clear classification, despite possible use against cables and pipelines. The register's focus on platforms carrying conventional weapons can miss drones used as ammunition or loitering munitions.

V. Under UN Security Council Resolution 1540

Resolution 1540 requires states to prevent non-state actors from acquiring weapons of mass destruction and their means of delivery. Its focus on industrial supply chains does not fully address agricultural drones that could disperse biological or chemical agents, or drones used for cross-border narcotics and illicit trade. The resolution should account for modern drone delivery capabilities and component monitoring.

VI. Under the Biological Weapons Convention (BWC, 1972)

The BWC addresses development, production and stockpiling of biological agents but does not clearly define unmanned aircraft as delivery platforms. Crop-dusting and vaccine-delivery systems use dispersal mechanisms that can be dual-use. A technical annex should define autonomous or remotely piloted systems capable of liquid dispersal as delivery platforms when exported to high-risk regions or non-state entities.

VII. Under UNCLOS

“In the territorial sea, submarines and other underwater vehicles are required to navigate on the surface and to show their flag.”

UNCLOS predates UUVs and does not explicitly address them. The surface-navigation rule is poorly suited to submerged autonomous operations, while definitions of ships and vessels leave uncertainty around unmanned systems and innocent passage.

“The following scenario illustrates a plausible risk under conditions of unregulated drone proliferation and the current governance gap.”

The Gdańsk Scenario

AI-generated illustrative scene of drones over a football stadium in Gdańsk
AI-generated image for illustration purposes

Imagine an evening in Gdańsk, near Poland's northern coast. A stadium is filled to its capacity of 42,615 for an international football match. A swarm of 20–30 legally purchased high-performance drones descends unnoticed. The first strike hits the pitch, killing players. As spectators flee through narrow exits, further drones target the crowd and turn the exits into danger zones.

Security forces are overwhelmed and a stampede causes further casualties. In the aftermath, the state is left without clear attribution or accountability. This hypothetical illustrates how a relatively small number of commercially available systems could create mass civilian harm without bringing down the stadium itself.

Towards a Drone Non-Proliferation Framework

Classical columns representing the proposed drone governance pillars
The proposed pillars (Source: iStock)

Gaps in current frameworks are not merely bureaucratic oversights; ambiguity can enable proliferation and sabotage. A Drone Non-Proliferation Framework (DNPF) should be grounded in the following pillars.

Pillar I — Advanced Export Control and Verification

Building on START I-style verification, a “stockpile-to-strike” tracking system would classify drones by operational capability rather than size or technical specification:

  1. Tier 1 — Unmanned Combat Aerial Vehicles (UCAVs)
  2. Tier 2 — Surveillance and reconnaissance systems
  3. Tier 3 — FPV and loitering munitions, including kamikaze systems
  4. Tier 4 — Logistical and heavy-lift drones

Mandatory on-site inspections would complement self-reporting under the ATT. A multilateral monitoring body could verify declared stockpiles and confirm that industrial dual-use systems have not been repurposed for combat.

Pillar II — Procedural Change in Air Operations

A standard notification protocol for drone operations in controlled or high-intensity airspace could reduce accidental escalation and protect civil aviation. Clear notice gives other states a signal of intent and helps distinguish civilian transit from operational zones. Civilian protection should be a procedural requirement under the framework.

Pillar III — Proliferation Containment and Mandatory KYC

Consumer drones exceeding defined performance thresholds should trigger Know Your Customer checks and national registration. High-performance systems and components should require import/export licences, similar to controls on small arms, to limit cross-border assembly from inexpensive open kits.

Pillar IV — Mandatory Integrated Geofencing

Flight controllers could connect to GNSS databases and restrict or land a drone when it enters designated sensitive areas, including critical infrastructure, military sites and borders. In conjunction with Pillar II, cellular connectivity could support landing procedures where a Notice to Airmen (NOTAM) is issued. Emergency remote landing could reduce risk, but geofencing is a barrier rather than a complete solution and requires safeguards against spoofing and misuse.

Pillar V — National Drone Registry and Multilateral Accountability

Each system would receive a unique identification number (UIN) and be tracked through its lifecycle. Ratifying states would submit annual stockpile summaries to a UN-led monitoring body and could share information on systems near sensitive borders to prevent accidental escalation. Failure to control exports or maintain a registry could trigger proportionate technology-related measures. A recovered drone's UIN could help investigators trace its origin and establish responsibility under international law.

Pillar VI — Freedom of the Seas

The framework should prohibit nuclear warheads and nuclear reactors on fully autonomous UUVs, given the risks of communication delays and malfunction underwater. To address UNCLOS Article 20, it could introduce a proposed standard for innocent passage, including resurfacing at least every 24 hours, display of a physical or digital flag of origin, and tamper-resistant tracking for vehicles operating near sensitive infrastructure. Such measures could support attribution after damage to an undersea cable.

Constraints

I. Major Power Compliance

States with advanced research programmes may resist stockpile disclosure and confidence-building measures, especially where autonomous systems are central to national strategy. This could limit adoption and ratification.

II. Technical Bypass Mechanisms

Actors may jailbreak flight controllers, alter source code or jam GNSS signals to evade geofencing and forced-landing functions. The framework would be only as reliable as its software, tamper resistance and fallback controls.

III. Uneven Implementation of NOTAMs

A party preparing a surprise strike may not issue a NOTAM, while false notices could be used to misdirect air defences. Procedures must protect civilian aviation without becoming an instrument of deception.

Strategic Implications for Poland and the EU

For Poland. Wider adoption of the DNPF could help counter hybrid tactics on the eastern border. After the September 2025 drone incursions, lack of attribution left room for deniable aggression. Under Pillar V, traceable systems could give Polish authorities verifiable evidence to present to NATO and the UN. Export controls and registration could also impede drone-facilitated cigarette and drug smuggling from Belarus.

Poland's energy and port infrastructure, including the Świnoujście LNG terminal, could benefit from geofenced protection zones. Pillar IV could reduce the risk of civilian or modified drones reaching sensitive assets, while recognizing the limitations of digital geofencing.

For the EU. The DNPF could standardize regulation across the internal market. Criminal groups use drones to move narcotics and contraband across borders; as maritime routes and underwater systems develop, these tools could also threaten civil infrastructure. NOTAMs and geofencing during periods of unrest may limit tactical surveillance of police while preserving transparent media access.

The Beginning of a Comprehensive System

The proposed pillars are interdependent: capability-based classification informs export controls, which support market regulation and verification. The DNPF does not aim to eliminate misuse, but to raise its cost, reduce anonymity and constrain escalation.

For Poland and the EU, such a framework addresses a practical security need as drones become tools of hybrid interference. The international community faces a choice: adapt governance to these technologies or allow existing gaps to widen. The DNPF offers a starting point.

Sources Consulted

  1. Dukowitz, Z. (2026, February 27). How much does a drone cost? Drone prices by type for 2026. Drone Pilot Ground School.
  2. How fast can a drone fly? Top speeds of various types. JOUAV (2025).
  3. Saini, R. K., Raju, M. S. V. K., & Chail, A. (2021). Cry in the sky. Industrial Psychiatry Journal, 30(Suppl 1), S15–S19.
  4. Hijazi, A. et al. (2017). Psychological dimensions of drone warfare.
  5. The Wassenaar Arrangement.
  6. Wassenaar Arrangement Secretariat (2026). List of Dual-Use Goods and Technologies and Munitions List.
  7. Peck, M. (2026, March 18). US Air Force special operations seeks kamikaze drones. Defense News.
  8. United Nations. The Arms Trade Treaty.
  9. Hecht, J. (2002, July 24). Fighter plane's laser may blind civilians. New Scientist.
  10. High Contracting Parties (1995). Protocol on Blinding Laser Weapons (Protocol IV).
  11. UN Register of Conventional Arms categories.
  12. UN Office for Disarmament Affairs. Biological Weapons Convention.
  13. United Nations Treaty Collection. Treaty record.
  14. United Nations Convention on the Law of the Sea.
  15. Polsat Plus Arena Gdańsk stadium profile. Transfermarkt.
  16. Pal, A. et al. (2022). Communication for Underwater Sensor Networks: A Comprehensive Summary. ACM Transactions on Sensor Networks, 19(1), 1–44.
  17. Adithya Suresh (2025, October 23). An assessment of Russian drone incursions on Polish soil and European/NATO airspace. Medium.
  18. PolskieRadio.pl. Smuggling drone carrying Belarusian cigarettes found near Poland’s eastern border.
  19. Europol (2026, January 27). Evolving tactics in maritime cocaine trafficking operations.

← Back to publicationsRead on Medium