Tokyo's ¥9.04 Trillion Pivot: How Japan's FY2026 Defense Budget Re-arms via Drones, AI and "Cross-Domain" Power
东京的9.04万亿拐点:日本2026财年防卫预算如何用无人机、AI与“跨域战力”重新武装
East Asia Military Watch · Institutional Briefing
东亚军事观察 · 怡心湖智库简报
0. Executive Summary / 核心判断
On 26 December 2025, the Japanese cabinet locked in the FY2026 defense budget at ¥9.0353 trillion (approx. $58 bn), up ¥334.9 bn (+9.4%) year-on-year and marking the 14th consecutive annual record. More important than the headline number is the line-item shift: Tokyo is no longer buying "defense in quantity"—it is buying offensive reach, unmanned attrition networks, and AI-enabled C2.
2025年12月26日,日本内阁敲定2026财年(2026.4–2027.3)防卫预算9.0353万亿日元,较上年增3349亿、增幅9.4%,连续14年创新高。比“破9万亿”更关键的是结构——日本不再按“专守防卫”买数量,而是在买防区外打击、无人消耗网、AI赋能指挥。
One-line take: FY2026 is the year Japan's budget stops pretending to be "reactive" and starts funding a deployable, intelligent, cross-domain strike force under the label of "dynamic deterrence."
一句话定性:2026财年是日本预算撕掉“被动反应”伪装、在“动态威慑”名下为“可部署、智能化、跨域打击力量”买单的拐点年。
1. The ¥9 Trillion Baseline: Scale, Trajectory, GDP Math
1. 9万亿基线:规模、轨迹与GDP账
|
Fiscal Year |
Defense Budget |
Note |
|---|---|---|
|
FY2022 |
¥5.4 tn |
pre–"new security documents" baseline |
|
FY2023 |
¥6.8 tn |
1st year of ¥43 tn / 5-yr plan |
|
FY2024 |
¥7.9 tn |
— |
|
FY2025 |
¥8.7005 tn |
— |
|
FY2026 |
¥9.0353 tn |
+9.4% YoY, 14th record, >2% GDP |
-
Under the 2022 "Security Three Documents", Tokyo pledged ¥43 tn over 2023–2027; FY2026 is year four, and the 2% of GDP target (NATO-style) is now reached two years ahead of the original 2030 schedule.
-
The five-year glide path points to an ~¥10 tn annual budget by FY2027, which would lock Japan into the world's third-largest defense spender behind the US and China by raw expenditure.
-
高市内阁把“GDP 2%”从2030年提前到2026年3月前达标,法理上把“专守防卫”的财政外壳拆了。
2. Where the Money Goes: From Steel to Software
2. 钱去哪了:从钢板到算法
Traditional platforms still get funded, but the marginal yen flows to standoff fires, unmanned systems, AI, space, cyber—the "new domains" (新域).
传统舰机仍拿钱,但边际日元流向防区外火力、无人系统、AI、太空、网络这五个“新域”。
2.1 Standoff Strike / 防区外打击(counter-base capability)
-
¥973.3 bn total for standoff fires; includes ¥177 bn for extended-range Type-12 (≥1000 km), ¥30.1 bn for hypersonics, ¥16 bn for sub-launched missiles.
-
First batch of extended Type-12 to deploy in Kumamoto, SW Japan, by March 2026—one year early.
2.2 Unmanned Systems / 无人作战:¥277.3 bn(≈2.5× FY2025)
-
SHIELD (盾牌) coastal layered network: ¥100.1 bn; procures ~10 types / several thousand UAS (incl. attack drones), USVs, UUVs; operational target March 2028; focus on SW Islands attrition belt.
-
Wide-area ISR UAS: ¥11.1 bn for 5 large UAS + ¥1.1 bn MQ-9B concept validation (23-aircraft buy later, deal value ~¥300 bn).
-
Loyal Wingman / GCAP: ¥4.8 bn for MHI stealth UCAV (interceptor + recon variants, ~10 m length, 200 kg payload to 200 km island). NATO-side GCAP trilateral (JP/UK/IT) gets >¥160 bn R&D, part of which funds the wingman software stack.
-
Total unmanned-related + missile coordination line: ¥312.8 bn.
2.3 AI as a Line Item / AI首次单列章节
FY2026 is the first budget with a dedicated "AI Utilization Promotion (AI活用推进)" subsection:
-
Target detection, intel fusion, C2 decision-aid
-
Human–machine teaming trials (有人—无人协同) with autonomous rules of engagement
-
Cyber Force expanded to ~3,310 personnel with AI-assisted decision support
2.4 Space & Org Restructure / 太空与编制
-
JASDF renamed "Air-Space Self-Defense Force (航空宇宙自卫队)"; new Space Operations Group (宇宙作战集团, ~880 personnel) under 3-star command; ¥1.1 bn for satellite-interference sensors.
-
Tactical AI satellite demo + next-gen defense tech demonstration satellite continued.
-
Ground: 15th Brigade (Naha) → Division, Pacific Defense Concept Office established.
3. Strategic Reading: Three Shifts That Matter
3. 战略解读:三个真正起变化的位移
(i) Reactive → Pre-emptive / 被动→先制
The ¥973.3 bn standoff line + submarine-launched cruise missiles + Type-12 ≥1000 km = a counter-base (对敌基地攻击) arsenal the 2018 NDPG would not have funded. "Dynamic deterrence" is now a procurement category, not a slogan.
(ii) Manned → Manned-Unmanned Teaming / 有人→有人无人混编
SHIELD is explicitly an asymmetric attrition belt: cheap drones/USVs exhaust high-value opponent munitions before a manned engagement. GCAP wingman extends the same logic to air combat. Japan accepts it is ~10–15 years behind the US in high-end UAVs, so it buys consumable mass + AI autonomy rather than chasing pure stealth parity.
(iii) Single-domain → Cross-domain / 单域→跨域
Air-Space ASDF + Space Ops Group + AI-C2 + cyber-AI + unmanned ISR = a joint kill-web concept borrowed from US CJADC2 but seated in Kanto. The budget is no longer "maritime defense of Japan"; it is "integrated deterrence from seabed to orbit."
4. Regional Fallout / 地区连锁反应
-
China: MOFA spokesperson already flagged the hike as "right-wing remilitarization"; PLAN/PLAAF will read SHIELD + Kumamoto Type-12 as a SW-Islands anti-access layer aimed at Taiwan contingency scenarios.
-
Korea: Tokyo's ¥9 tn / ~2% GDP vs. Seoul's ~2.5% but smaller absolute base reshapes bilateral defense-industrial competition in shipbuilding & UAV supply chains.
-
US: Washington gets its 2% ask two years early, but also a partner building independent standoff fires + indigenous GCAP—alliance plus autonomy at once.
-
Constitutional fact: Article 9 is untouched on paper; on budget lines it is now financing hypersonics, attack drones, and a space force. The gap between text and expenditure is the real story.
5. Watchlist to 2027 / 到2027年的观察清单
-
March 2026 – first extended-Type-12 battery live in Kumamoto.
-
March 2028 – SHIELD first operational node; track UAS monthly burn rate.
-
FY2027 budget request – does Tokyo cross ¥10 tn? If yes, 2% is structural, not cyclical.
-
GCAP wingman – first manned-unmanned hookup flight (MHI timeline).
-
Space Ops Group – first published tasking order covering LEO SDA (space domain awareness).
Bottom line for policymakers
FY2026 is not "more of the same." It is the budgetary instantiation of Japan's exit from post-war defensive restraint—executed through drones, algorithms, and standoff missiles rather than battleships. The 9-trillion figure is the headline; the SHIELD + AI-C2 + GCAP wingman + Air-Space ASDF quartet is the architecture. Any East Asia threat assessment that still models Japan as a "host-nation logistics hub" is now mis-modeled.
给决策者的底线
2026财年不是“又涨一点”。它是日本退出战后防御克制在预算层面的实例化——借无人机、算法和防区外导弹完成,而不是借战列舰。9万亿是标题,SHIELD+AI-C2+GCAP僚机+航空宇宙自卫队这四件才是架构。凡仍把日本建模为“东道国后勤枢纽”的东亚威胁评估,都已失真。
Sources: Japan MoF/MoD FY2026 budget release; Kyodo/AP/Jiji 26 Dec 2025; 中国军网 2026-07-20; 环球时报 2025-12-26; 上海日本研究交流中心 2025-12-29. Figures rounded to stated precision.
资料源:日本财务省/防卫省2026财年预算文件;共同社/美联社/时事社2025-12-26;中国军网2026-07-20;环球时报2025-12-26;上海日本研究交流中心2025-12-29。数值按原文精度取整。
此文由 怡心湖 编辑,若您觉得有益,欢迎分享转发!:首页 > 观·世界 » 东京的9.04万亿拐点:日本2026财年防卫预算如何用无人机、AI与“跨域战力”重新武装
The Boundary of Mind: Navigating
数字提督:波音防务人工智能全面融入
The Cognitive Cockpit: AI Algori
The Silent Storm: 无人机蜂群、机
AI 时代的 Linux 系统管理:从基础到
【怡心湖智库前瞻】Kimi K3与AI算
【智库观察】Kimi K3 Moment:当开源