Sir Airffic GoWhich of these five is a tailwind for your company, and which a headwind? Saying that clearly is step one of strategy.
Map of the five forces
- US–China decoupling (DJI curbs, NDAA)
- War & defense demand (Ukraine)
- Regulatory unlock (BVLOS, Part 108, Level 4)
- Edge AI chips
- Foundation models & generative AI
Force 1: US–China decoupling
In 2020 the US added DJI to its export-control Entity List. Defense budget laws (NDAA) and other measures have since restricted government purchases of Chinese drones.
In the US and allied countries, not being made in China became a selling point. Being on the Pentagon-vetted Blue UAS list is a sales weapon.
This matters for Japan too: security and infrastructure buyers increasingly ask for domestic or transparent supply chains.
Force 2: war and defense demand
In the war in Ukraine that began in 2022, cheap small drones changed how wars are fought. Jamming became routine, and demand for drones that fly to targets on their own surged.
In 2023 the US launched Replicator, a plan to field large numbers of cheap autonomous systems. Large sums also flowed to European defense drone start-ups.
Defense money rapidly advances autonomy, sensing and counter-drone technology. Respecting export controls and ethical lines is a condition for lasting.
Sir Airffic GoThe defense tailwind is strong. But draw the line on what you will and won't do first, or you can lose trust overnight.
Force 3: regulatory opening
When BVLOS becomes routine, one pilot can run many aircraft and the economics of delivery and inspection change.
Japan's Level 4 (2022), Europe's U-space and the US Part 108 proposal (2025) all open BVLOS on condition of aircraft safety and airspace management.
Regulatory opening therefore creates demand for airspace OS and operations management in itself.
Force 4: edge AI chips
Power-efficient AI chips built for phones and cars now fit on drones.
Self-localisation and object detection that were lab-only ten years ago are now routine on commercial aircraft.
Better chips make brain functions commonplace. Differentiation gradually moves toward data and operations.
Force 5: foundation models and generative AI
Large models that understand images and language together are starting to be used in robots and drones.
As they mature, aircraft may carry out a mission from a simple instruction such as "inspect every defect on this bridge" without detailed programming.
Systems that guarantee sky safety have not caught up. Recording, verifying and coordinating AI decisions across the airspace is emerging as new demand.
Who each force helps
| Force | Tailwind for | Headwind for |
|---|---|---|
| US–China decoupling | US/allied and domestic aircraft makers | Operators reliant on Chinese aircraft |
| Defense demand | Autonomy software, C-UAS, defense drone start-ups | Small civil-only firms (talent and capital drawn away) |
| Regulatory opening | Operations, airspace OS, DAA, remote-operations firms | Services reliant on manual operations |
| Edge AI chips | Chipmakers, software differentiators | Firms selling perception alone (commoditised) |
| Foundation models | Firms owning data and verification systems | Firms hand-building narrow AI |
This mapping is the editors' interpretation of public information; individual companies may differ.
Frequently asked questions
Why are DJI drones restricted in the United States?
Citing national-security concerns, the US government added DJI to the export-control Entity List and has used defense budget laws and other measures to restrict government purchases of Chinese drones. DJI disputes these concerns.
Which businesses grow when drone rules loosen?
When BVLOS becomes routine, delivery, wide-area inspection and emergency response can be run with one person overseeing many aircraft, increasing demand for operations and airspace management services.