Taiwan Solved Its Drone Funding Problem. Ownership Risk Is the One That's Left

Taiwan's drone industry no longer has a money problem. The Executive Yuan's September 3 supplementary allocation put NT$55.9 billion behind more than 40,000 coastal attack drones, on top of the six-year, NT$240 billion uncrewed-vehicle act the Legislative Yuan passed a week earlier, according to SET News's reporting on the budget. What the industry has instead is a harder question, and it arrived the same season as the money: can the companies meant to build all of this survive the political pressure that comes with being chosen to do it.

Start with the scale of what just got funded, because it changes the shape of every other question in this piece. Of the September 3 allocation, NT$13.8 billion is earmarked for this fiscal year and NT$42.1 billion for next, on top of the NT$240 billion Legislative Yuan act, funding more than 40,000 coastal attack drones, roughly 600 coastal surveillance drones, and over 100 small unmanned surface vessels. The same week, Section 232 tariffs on drones and drone components took effect in the US market, with Taiwan-origin products eligible for a maximum 15 percent rate against tariffs of up to 100 percent on Chinese-linked supply. Domestic demand and an external tariff wall are now pulling in the same direction at the same time, a combination the industry hasn't had before.

DSET's newly published assessment of the Lai administration's drone strategy through 2024–2026 puts scale on that pull: military and civilian UAV procurement went from 3,422 units in 2024 to 48,750 in 2025, and the Ministry of National Defense has now proposed requirements for roughly 210,000 coastal-surveillance and coastal-attack units, alongside 1,320 one-way-attack USVs. Exports climbed from 3,472 units worth NT$140 million in 2024 to 122,726 units worth NT$2.95 billion in 2025, with cumulative 2026 exports already at 191,373 units by the end of the second quarter, per DSET's report. That's the growth curve the industry is now supposed to fill.

One recent case shows why funding alone doesn't settle the question. GEOSAT (經緯航太), a Taiwan drone maker that won a NT$7 billion military tender in 2023 and took on Wistron subsidiary Dingchuang as its controlling shareholder in 2024, was named by China's Commerce Ministry to a dual-use export control list in July 2025. Wistron has since cut GEOSAT's capital by roughly 70 percent and merged it into a newly created subsidiary, Xiangyu Technology (翔譽科技股份有限公司), according to the merger filing on the Taiwan Stock Exchange market observation system; GEOSAT no longer exists as a standalone company. One commentator has argued this amounts to Beijing sidelining a flagship supplier through pressure on its parent rather than the company itself, a reading Wistron, GEOSAT, and the government haven't confirmed.

What the audit trail says is still open

DSET's report is useful precisely because it doesn't read as a victory lap. It credits Thunder Tiger's entry onto the US Blue UAS Cleared List, ITRI's authorization as a Green UAS cybersecurity assessor, and CMMC Level 2 certification at both AIDC and NCSIST, but it's equally direct about what hasn't caught up: Taiwan still lacks mutual-recognition mechanisms with the US, Japan, and Europe, existing verification systems remain incomplete for continuous tracking of bills of materials, component provenance, firmware updates, and supplier ownership changes, and the "three chips and two software" localization push, backed by just NT$326 million across seven approved projects, leaves communications, navigation, autonomous control, thermal imaging, and mission software still dependent on foreign suppliers. Upstream, battery cells, cathode and anode materials, and rare-earth magnets remain PRC-constrained. DSET's own recommendation is to extend non-PRC sourcing requirements from the component level to upstream materials and to move procurement onto a three-to-five-year rolling basis rather than year-to-year appropriation, precisely because annual budget review, however generously funded, doesn't give a supplier the multi-year visibility it needs to plan capacity.

Ownership structure sits outside that audit scope entirely. A bill-of-materials review checks what's inside a product, not who owns the company selling it or how much unrelated commercial exposure that owner is carrying elsewhere. A clean BOM would have told a buyer nothing about GEOSAT's real exposure, because the risk sat one level up, in who owned the company, not in the parts list. That's a gap Vietnam's own suppliers have had to think about too, for different reasons.

What Gremsy did differently

Vietnam's Gremsy offers a contrast on exactly that structural point. Gremsy, a stabilized-gimbal and payload maker based in Ho Chi Minh City's Saigon Hi-Tech Park, cleared three payloads, the ORUS L NDAA, LYNX NDAA, and VIO F1, onto the Blue UAS Cleared List in August, becoming the first Vietnamese company and the only Asia-Pacific payload maker on the list. It did this through Gremsy USA, a separate US-incorporated entity with its own CEO, Ryan Le, which the company says engineers and builds the compliant hardware domestically rather than shipping finished Vietnamese assemblies for approval, according to DroneLife's coverage of the launch. The Vietnamese parent, Gremsy JSC, runs the rest of the business. The cleared products are also built around Teledyne FLIR's NDAA-compliant thermal cores rather than a Gremsy-developed sensor, a subsystem rather than a whole-aircraft claim.

Zhou Yu-ping makes a related argument about where Taiwan's drone industry should be headed: that a supply chain built only to be non-Chinese is an advantage granted by someone else's policy choice, and should evolve into a trusted and eventually an indispensable one, built around specific subsystems, rather than a single whole-aircraft brand riding on one parent company's risk tolerance.

What we're watching

The most concrete near-term test is whether the NT$55.9 billion supplementary allocation clears the Legislative Yuan intact, since a National Party spokesperson has already argued publicly that the sum validates their preference for annual rather than special-act budgeting, a dispute that has repeatedly slowed unmanned-vehicle appropriations this year. We're also watching whether any of GEOSAT's former contracts or technical capability resurface under Xiangyu Technology, and whether any Taiwan supplier follows the Gremsy model, a separate US-incorporated entity holding US clearances, rather than a single-entity structure. And we're watching whether DSET's recommendation on multi-year rolling procurement and upstream disclosure gets taken up, since that's the mechanism most directly aimed at the ownership-risk gap a bill-of-materials audit doesn't cover.

→ Subscribe to Taiwan Drone Weekly for ongoing coverage: www.taiwandrones.com/subscribe
→ Suppliers: contact us at info@taiwandrones.com

This analysis draws on public reporting from TechNews, SET News, CNA, DroneLife, and Unmanned Systems Technology, on DSET's September 2026 research report on Taiwan's UAV industry, on a commentary published by UP Media, on Taiwan Stock Exchange market observation system filings related to the Wistron-GEOSAT-Xiangyu merger, on Gremsy's and Gremsy USA's own published statements about their Blue UAS clearance, and on TaiwanDrones' own prior reporting. The characterization of Wistron's motives as a deliberate response to Chinese pressure reflects one commentator's published opinion, not a confirmed corporate or government account, and is identified as such. It reflects the state of public information as of September 24, 2026 and should not be read as investment advice or an endorsement of any company or product.