Japan’s Drone Budget Creates a Contest Between Giants and Startups

A defense budget begins to create a market

Japan’s drone industry was long centered on surveying, inspection, agriculture and disaster response. Defense applications existed, but the idea of networks of expendable small unmanned systems was not central to procurement.

That changed in 2026. A rapidly expanding unmanned-systems budget now covers SHIELD, small attack UAVs, ship-launched aircraft, interceptors, unmanned surface vessels and underwater systems.

The competitors include Mitsubishi Heavy, Kawasaki Heavy, IHI, NEC and Mitsubishi Electric, but also domestic startups such as ACSL, Liberaware, Terra Drone, Blue Innovation and Prodrone. Foreign firms including Anduril, Shield AI and General Atomics are also approaching the Japanese market.

The budget is not only buying aircraft. It could shift Japan’s defense industry from long-cycle, low-volume platforms toward a software-intensive market with faster product turnover and more frequent company entry.

The scale of fiscal 2026 spending

Japan’s fiscal 2026 budget allocates ¥100.1 billion for unmanned assets associated with SHIELD, including modular UAVs, several types of small attack drones, anti-ship systems and UAVs for protecting radar sites.

The Financial Times reported total unmanned-vehicle procurement of about ¥277 billion, up from ¥111 billion the previous year. Japan has also reportedly set a goal of annual production capacity of 80,000 drones by 2030.

Large budgets allow companies to hire, build factories and invest in testing. But if orders fall sharply after a few years, startups can be crushed by fixed costs. Market formation requires predictable multi-year demand.

The strengths of the traditional primes

Mitsubishi Heavy and other large contractors do more than build airframes. They integrate missiles, radar, ships, aircraft, satellites, communications and command systems.

A defense drone must receive sensor data, identify targets, coordinate with aircraft and ships and continue operating under electronic attack. Large companies have advantages in security, certification, maintenance and long-term contracting.

MHI used Shield AI’s Hivemind Enterprise environment to move from autonomy development to UAV flight demonstration in eight weeks, showing that a traditional prime can adopt software-centered speed.

Still, large contractors can be slow and expensive, with requirements frozen before development. That process may be too heavy for tactical drones that change every few months.

What startups bring

Startups specialize in rapid product cycles and narrow technical problems: small aircraft, flight control, swarming, image analysis, indoor navigation, communications and fleet management.

ACSL emphasizes secure Japanese-made drones. Liberaware builds very small aircraft for confined spaces. Blue Innovation integrates fleets of drones and robots. Prodrone focuses on larger industrial platforms, while Terra Drone combines operations, inspection and overseas expansion.

These companies are not purely military. Civilian markets generate volume, data and operational experience. Defense demand can accelerate development and production.

But defense contracts impose cost accounting, security rules, export controls and long-term support requirements. Administrative compliance can become more difficult than engineering.

Foreign companies enter the contest

Anduril has reportedly discussed producing drones at Nissan’s Oppama plant. Shield AI has supplied autonomy tools to MHI. Large reconnaissance programs involve companies such as General Atomics.

Foreign firms bring battlefield data, private capital and mature software. Japan can acquire capability faster than if every component is developed domestically.

Dependence creates risks involving source code, data, update rights, export licenses and spare parts. A crisis may expose limits imposed by a foreign government.

Postwar procurement favored established contractors

Japan’s defense market developed around stable relationships between MOD and large industrial groups. Platforms were expensive, low-volume and sustained for decades. Trust and quality mattered, but entry was difficult.

Arms-export restrictions left the Self-Defense Forces as almost the only customer. Losing one contract could mean losing the entire market. Supply-base preservation often mattered more than open competition.

That structure can be reasonable for submarines and fighters. It is less suited to drones and software whose underlying civilian technologies change rapidly.

Why startups struggle to enter defense

Defense startups face a classic valley of death. A prototype may be ready years before formal requirements and procurement. Even after selection, production capital is needed before revenue arrives.

MOD contracting requires qualifications, audits, security and quality systems that small firms often lack. Classified work can complicate foreign investment and international hiring.

Japan’s venture-capital market is smaller than America’s, and some ESG policies avoid weapons. A startup dependent on one defense order can fail when the budget moves.

Can the 2023 industrial-base law change entry?

The 2023 defense-production law created support for supply-chain resilience, factory efficiency, cybersecurity and succession.

Japan also seeks dual-use SMEs and university technology through research institutes and technology solicitations.

But if support only preserves subcontractors under incumbent primes, independent challengers will not emerge. Startups need small direct contracts and rapid review mechanisms that allow them to act as prime suppliers.

The Defense Innovation Unit model

The United States created the Defense Innovation Unit in 2015 to connect technology companies with the Pentagon. Its Commercial Solutions Opening process can test civilian products in months and move successful systems toward production.

Anduril, Shield AI and Palantir grew while competing with established primes. Private capital often builds products before government becomes the customer.

Japan cannot copy the model exactly because its market, venture capital and export environment differ. But rapid field trials, small contracts that lead directly to production and multi-year orders are relevant lessons.

Ukraine’s frontline procurement

Ukraine connected soldiers, volunteers, startups and government platforms in short development cycles. Frontline feedback could become a design change within weeks.

The model is fast but inconsistent in quality, compatibility, security and parts control. Japan cannot simply import wartime emergency procurement.

The lesson is to reduce distance between user and developer, share failures quickly and avoid freezing requirements too early.

Three ways to allocate the budget

Prime-led programs clarify responsibility but can become slow and costly. Competitive trials encourage innovation but duplicate spending. Open architecture improves competition and upgradeability but creates integration and cyber challenges.

A Japanese hybrid may place system responsibility with a prime while allowing startups to compete at the module level.

Why price cannot be the only criterion

Cheap drones matter, but lowest cost alone is dangerous. A platform that can be hijacked, jammed, cut off from parts or left without software updates may fail in combat.

Procurement should evaluate jamming resistance, update speed, domestic repair, component traceability, data rights and cybersecurity.

Requirements must still be tiered. Expendable, reusable and high-assurance systems should not all be forced into the same certification standard.

Who owns the data and software?

The most valuable assets may be flight data, sensor records, failures and AI models rather than the airframe.

If government buys software but the vendor controls all learning data and updates, switching becomes difficult. If government demands all intellectual property, private investment may fall.

Contracts need clear rules on operational data, model improvement, source-code escrow and emergency update rights.

Domestic production and economic security

Japan sees dependence on Chinese-made drones as a security risk, particularly where government and infrastructure imagery or location data may leave the country.

Domestic production improves trust and supply resilience but usually costs more. Full national self-sufficiency in motors, batteries, chips and cameras is unrealistic.

A practical strategy combines multiple suppliers, trusted-country partnerships, domestic final integration and software audit.

Exports can enlarge the market

Japan’s 2026 relaxation of defense-export rules could permit case-by-case sales of combat drones and other systems.

Exports would expand production beyond the small domestic market and give startups recurring revenue. Demand exists in Southeast Asia, Europe, Australia and the Middle East for surveillance, maritime security, disaster response and unmanned systems.

Japan must still manage human-rights risk, conflict use and third-country transfer. Industrial policy cannot be separated from diplomacy and ethics.

The field should decide the winners

Drone companies should ultimately be judged in demanding field tests, not at trade shows. Wind, rain, salt, electronic warfare, communications loss, night operations and shipboard launch should be measured.

Units need a way to use prototypes quickly and return direct feedback. Failure should become learning rather than something hidden as a contract embarrassment.

Large-company integration, startup speed and foreign battlefield experience should be compared in the same operational environment.

Japan.co.jp view: fund an ecosystem, not a list of companies

The larger Japan’s drone budget becomes, the greater the danger that money simply flows to established primes while startups remain permanent subcontractors.

The opposite mistake is romanticizing startups and ignoring maintenance, security and integration. Defense systems cannot fail like consumer apps and wait for tomorrow’s update.

Japan needs an ecosystem in which several teams compete by mission, connect through common standards and move winning technology rapidly into production.

Success should not be measured by how many firms receive subsidies. It should be measured by whether cheaper, faster and safer capability reaches units and can be improved continuously inside Japan.

The fiscal 2026 budget is the first major opportunity to create a Japanese defense-startup market—and a test of whether old procurement culture will absorb or enable new technology.