Hello, this is Ryuta Hamamoto from TIMEWELL.
When Japan's Cabinet approved the Growth Strategy on 21 July 2026, the coverage fixed on one number: ¥370 trillion. I started reading with the same impression, then opened the Public-Private Investment Roadmap published by the Cabinet Secretariat and found the picture was messier than the headline. Two of the 62 funded items have no figure at all. One has an investment number larger than its economic ripple effect. And the strategy itself carries a note saying the per-item ripple effects cannot be added together.
So this is not a document you read by summing things up. What interested me was not the size of the number but two other things: which sectors the government was willing to name, and how it intends to change the way budgets are built. This piece walks through both, using the 399 pages of primary documents rather than the coverage.
If you work in export control or counterparty screening, it is worth checking where your company sits within those 62 items. Our three-minute export compliance check is a quick way to see where you currently stand before the second half of this article.
What the Growth Strategy actually decided
The Growth Strategy is a Cabinet decision dated 21 July 20261. On the same day, the Japan Growth Strategy Headquarters, meeting for the second time by written procedure, adopted the Public-Private Investment Roadmap for the major products and technologies within the 17 strategic sectors2. The two are meant to be read together. The strategy explains what was chosen and why; the roadmap sets out how much money and which support measures, five pages per item.
The process was quick. The Headquarters first met on 4 November 2025, and the Japan Growth Strategy Council met seven times between 10 November 2025 and 21 July 2026. Under nine months to produce 399 pages of policy.
The structure rests on splitting investment in two. The first is "risk management investment", aimed at economic security, food security, energy and resource security, health security, national resilience and cybersecurity. The second is "growth investment", aimed at bringing frontier technology to market. The first secures what the document calls Japan's autonomy; the second, its indispensability. Accelerating both is AI transformation, which the strategy treats as a cross-cutting engine rather than a sector of its own.
The diagnosis behind all this is unusually blunt for a policy paper. "Japan's potential growth rate lags other major advanced economies, but our indicators for innovation capacity and labour efficiency are no worse than theirs. What is overwhelmingly missing is domestic investment." Whether that diagnosis is right is arguable. As a piece of official writing, it is refreshingly direct.
Alongside the 17 sectors, the strategy sets out eight cross-cutting themes covering startups, talent, finance and more. I covered the startup side of that in an earlier piece on the Startup Package, so this article stays with the 17 sectors and the 62 items.
The criteria for picking the 62 items are stated too: the need to reduce domestic economic security and other risks; the possibility of capturing overseas markets; and the novelty of the underlying technology. Economic security comes first in that list, and I do not think the ordering is accidental.
The full list of 17 sectors and 62 items
Figures are in trillions of yen. Where no year is noted, the horizon is cumulative through fiscal 2040.
(1) AI and semiconductors (sector total 101.6)
| Item | Investment | Ripple effect |
|---|---|---|
| Physical AI (AI robots in particular) | 10.5 | 144.4 |
| Semiconductors at the core of physical intelligent systems | 68.0 | 443.3 |
| Vertical AI (domain-specific AI) | 23.1 | 222.0 |
(2) Digital and cybersecurity (sector total 55.4)
| Item | Investment | Ripple effect |
|---|---|---|
| Data platforms (to FY2035) | 0.9 | 2.2 |
| Secure AX/DX infrastructure for national and local government (to FY2035) | 7.4 | 10.9 |
| Advanced security products and services for the AI era (to FY2035) | 1.0 | 3.3 |
| Cloud, data centres and storage batteries (to FY2035) | 32.7 | 107.1 |
| Cloud-native healthcare DX infrastructure | 5.2 | 15.2 |
| Autonomous driving technology | 8.2 | 187.3 |
(3) Information and communications (sector total 28.8)
| Item | Investment | Ripple effect |
|---|---|---|
| All-photonics network (APN) | 5.9 | 36.2 |
| Submarine cables | 2.4 | 16.2 |
| Next-generation wireless (NTN, 5G/Beyond 5G) | 20.5 | 223.5 |
(4) Quantum (sector total 13.2)
| Item | Investment | Ripple effect |
|---|---|---|
| Quantum computing | 10.3 | 127.5 |
| Quantum communications and networks | 1.5 | 8.5 |
| Quantum sensing | 1.4 | 12.5 |
(5) Defence industry
| Item | Investment | Ripple effect |
|---|---|---|
| Small unmanned aircraft | 0.4 | 5.6 |
| Naval vessels | approx. ¥340bn (FY2026 budget) | under review |
| Dual-use technology | 4.3 | 117.7 |
(6) Aviation and space (sector total 18.5)
| Item | Investment | Ripple effect |
|---|---|---|
| Commercial aircraft (next single-aisle and next-generation) | 3.5 | 16.1 |
| Unmanned aircraft | 0.3 | 5.6 |
| Flying cars | 0.4 | 1.9 |
| Rockets and launch sites | 2.3 | 5.5 |
| Satellites and satellite services | 6.4 | 30.6 |
| Lunar exploration and low-orbit technology | 5.6 | 24.1 |
(7) Ocean (sector total 3.3)
| Item | Investment | Ripple effect |
|---|---|---|
| Marine unmanned vehicles (ocean drones) | 1.2 | 9.4 |
| Maritime domain awareness (MDA) | 1.2 | 8.7 |
| Innovative seabed development technology | 0.9 | 2.2 |
(8) Shipbuilding
| Item | Investment | Ripple effect |
|---|---|---|
| Next-generation vessels (to FY2034) | 1.0 | 9.7 |
| Ship repair (to FY2035) | 0.1 | 0.3 |
| LNG carriers | to be determined | to be determined |
(9) Materials: critical minerals and components (sector total 16.9)
| Item | Investment | Ripple effect |
|---|---|---|
| Permanent magnets | 0.2 | 3.2 |
| Green steel | 4.2 | 10.4 |
| Innovative metal components and materials | 0.3 | 3.3 |
| Low-carbon metal components (non-steel) | 0.7 | 6.7 |
| Smelting, separation and sorting from primary and secondary raw materials | 6.3 | 1.1 |
| AI-enabled composite new materials | 5.2 | 14.4 |
(10) Synthetic biology and bio (sector total 33.6)
| Item | Investment | Ripple effect |
|---|---|---|
| Bio-manufacturing | 12.8 | 66.7 |
| Biopharmaceuticals and regenerative medicine products | 20.8 | 174.9 |
(11) Drug discovery and advanced medicine (sector total 64.1)
| Item | Investment | Ripple effect |
|---|---|---|
| First-in-class and best-in-class products | 23.4 | 162.1 |
| Infectious disease countermeasure products | 7.2 | 29.2 |
| Biopharmaceuticals and regenerative medicine (restated) | 20.8 | 174.9 |
| Advanced medicine using innovative devices | 11.6 | 105.8 |
| Healthcare services using life-log data | 1.1 | 17.3 |
(12) Resource and energy security, GX (sector total 28.8)
| Item | Investment | Ripple effect |
|---|---|---|
| Next-generation solar cells (perovskite and similar) | 4.1 | 22.9 |
| Hydrogen and related fuels | 6.2 | 85.3 |
| Green steel (restated) | 4.2 | 10.4 |
| Next-generation geothermal | 1.0 | 2.8 |
| Offshore wind | 5.1 | 17.3 |
| Next-generation advanced reactors | 5.0 | 11.1 |
| GX chemicals | 3.2 | 75.5 |
(13) to (17) Remaining sectors
| Sector | Item | Investment | Ripple effect |
|---|---|---|---|
| Fusion energy | Fusion energy | 3.1 | 9.4 |
| Disaster prevention and resilience | Disaster prevention technology (to FY2030) | 2.6 | 14.5 |
| Port logistics | Cargo handling machinery | 0.4 | 1.0 |
| Port logistics | Cyber Port (port logistics DX) | 0.2 | 0.4 |
| Port logistics | Next-generation warehouses | 0.6 | 1.4 |
| Food tech | Plant factories | 4.6 | 36.8 |
| Food tech | Land-based aquaculture | 2.9 | 47.1 |
| Food tech | Food machinery | 1.2 | 18.8 |
| Food tech | Novel foods | 1.0 | 18.8 |
| Content | Games (to FY2033) | 24.5 | 191.4 |
| Content | Anime (to FY2033) | 3.3 | 93.5 |
| Content | Manga (to FY2033) | 1.6 | 17.2 |
| Content | Music (to FY2033) | 3.0 | 15.0 |
| Content | Live-action film (to FY2033) | 1.3 | 8.9 |
Laid out like this, the character of the strategy becomes visible. Semiconductors take the largest single allocation at 68.0 trillion, roughly double the 32.7 trillion for cloud, data centres and batteries. The physical AI page adds a note that investment in AI infrastructure such as data centres is counted inside the cloud and data centre line, which tells you the compute build-out has deliberately been spread across several items.
The other allocation that caught my eye was games. A single content category carries 24.5 trillion, more than the entire aviation and space sector at 18.5 trillion. Add anime, manga, music and live-action and the content sector reaches 33.7 trillion, almost exactly level with synthetic biology and bio at 33.6. Placing cultural industries inside the same framework as security and advanced medicine is a bolder call than it first appears.
Replace siloed classification work with AI.
METI's FY2024 data shows 52% of foreign exchange law violations stem from classification errors. Download the TRAFEED product catalog covering features and rollout.
How to read the ¥370 trillion figure
Here is where the coverage went astray.
First, ¥370 trillion is not government spending. The text reads: "domestic investment, public and private, expected across the 62 major products and technologies in the 17 strategic sectors amounts to a cumulative total of over ¥370 trillion by fiscal 2040." Public and private combined. Read it as a fiscal commitment and everything downstream goes wrong.
Second, the ripple effects cannot be summed. Note 2 of the roadmap is explicit: "there is overlap in the economic ripple effects between the major products and technologies, and the item-by-item figures cannot be aggregated." Lining up 443.3 for semiconductors, 223.5 for next-generation wireless and 222.0 for vertical AI and reaching for a calculator is a reading the document itself forbids.
Third, two of the 62 items carry no figure. For naval vessels, the investment line says only "investment including defence procurement in the naval vessel field is approximately ¥340 billion (FY2026 budget)", a single-year number, and the ripple effect reads "to be considered in light of the discussions on revising the three security documents". In other words, not yet calculated. With the revision of Japan's three strategic documents due before year-end, the number could not be fixed in time. LNG carriers are blanker still: both figures read "to be examined further as discussions among the parties continue". The 62 items were not scrutinised to a uniform depth. I wrote about the security document review itself in an earlier piece on the expert panel.
Fourth, one item shows investment exceeding ripple effect. In the materials sector, smelting and sorting from primary and secondary raw materials carries 6.3 trillion of investment against 1.1 trillion of ripple effect. That inversion is not a typo. The same page notes that the investment figure includes overseas mine development and smelting, while the ripple-effect calculation excludes overseas investment. Money that leaves the country is not counted as a domestic effect, which is the correct treatment. Without reading the footnote, though, you would mark it down as a poor-return item.
Fifth, two items are counted twice. Green steel appears under both materials and resource and energy security; biopharmaceuticals and regenerative medicine appears under both synthetic biology and drug discovery. The roadmap contains 64 blocks; subtract the two restatements and you have 62.
One more pair of numbers gets confused regularly. The strategy says that by fiscal 2040, "domestic private capital investment reaches ¥230 trillion a year and GDP approaches ¥1,100 trillion" as a projection, while Chapter IV positions "¥250 trillion in fiscal 2040" as the new public-private target for the same measure. 230 is the estimate, 250 is the goal. Different animals, so quote them separately.
"Risk management investment" is economic security by another name
Now the part that matters most for anyone doing export control or supplier management.
Reading the whole document, my strongest impression was that the new phrase "risk management investment" functions as a synonym for economic security. Of the five drafting principles for the roadmap, the third reads: "achieve Japan's autonomy and indispensability from the standpoint of economic security." The first criterion for selecting the 62 items is economic security as well. This is an industrial policy paper whose underlying motive is supply chain security.
One passage makes it concrete. On permanent magnets: "with supply destabilised by tightened export controls in a particular country, securing autonomy and indispensability is urgent." The phrase "a particular country" is the government's own wording, and the original goes no further. I am not going to fill in the name here either. What matters is the structural fact that another country's export control measures are now written into Japanese industrial policy as a stated premise. Export control has stopped being purely a question of paperwork for exporters. It appears in national strategy as a procurement-side risk.
The same pattern shows up elsewhere. The materials sector is built around autonomy in critical minerals and components. All three ocean items connect explicitly to security: the marine unmanned vehicle page notes "growing importance of unmanned assets in the security field and as dual-use technology", and the defence sector carries dual-use technology as a standalone funded item at 4.3 trillion of investment and 117.7 trillion of ripple effect.
That last point lands directly on practitioners. When money flows into dual-use development, companies that never had to think about classification start operating in territory where it applies. Sensors, motors, reduction gears, metal components. Civilian on their own, potentially controlled depending on what they end up inside. Winning a subsidy and then discovering you need an export control programme is a common enough sequence, and it is the wrong way round.
We work on exactly this problem: TRAFEED is an AI export control agent that supports classification against Japan's METI criteria and screens counterparties. Final classification remains the responsibility of your own export control officer. But before any tool, the useful first move is simply establishing where your company sits among the 62 items, because that tells you which rules are heading your way.
For adjacent policy, the 2026 fiscal policy and defence build-up and the Aerospace Self-Defense Force reorganisation cover the same current from different angles.
The budget rules changed more than the numbers did
The coverage chased the money. The more consequential change, to my mind, sits in Chapter IV, where the mechanics of budgeting were rewritten.
At the centre is a new "Strong and Prosperous Japan Investment Framework", with four features.
The first is the removal of request ceilings. The text: "the framework will not set request ceilings, so that necessary amounts, including unpriced item requests, can be requested appropriately." No cap at the budget request stage.
The second is special account ring-fencing and bridge bonds. "For fields of particular importance to economic security, funds will be managed separately in a special account, sufficient financing will be secured across multiple years, and adequate scale will be assured through the issuance of bridge government bonds backed by redemption sources." Money placed outside the single-year budget cycle.
The third is multi-year planning as the default, and the strategy explains why: "to improve predictability for the private sector." From a company's side this may be the most valuable change of the four. Capital investment decisions run on multi-year horizons, and subsidies that expire annually have always been awkward to build around.
The fourth is a commitment to cut what does not work: "progress against those plans will be reviewed regularly, and budgets with weak investment-inducing effects will be revised flexibly." A PDCA mechanism runs under the Growth Strategy Council.
Honestly, the fourth is the one I will be watching. Removing ceilings, ring-fencing accounts and securing multi-year funding all make it easier to spend. Easier spending without a working brake produces rigid budgets rather than good ones. The phrase "this must not end up as a picture of a rice cake" appears twice in the original, which is a Japanese way of saying "empty promise" and suggests the drafters know the history.
One further detail worth flagging is anchor tenancy. Both marine unmanned vehicles and MDA commit to "establishing a system to promote market formation and expansion through anchor tenancy", and the defence sector lists "creating leading demand through public procurement in government demand areas" as investment content. The government becomes the first customer rather than merely subsidising development costs. If that is genuinely carried out, it means far more to a startup than a grant does.
What to do with this document
Three things, which are what I did myself.
Locate your company among the 62 items. Not being a finished-goods maker does not put you outside this. The roadmap writes investment content down to the component level, naming "R&D and capital investment by suppliers of critical components such as reduction gears, motors, sensors and storage batteries". If you do not know which value chain you sit in, you have no basis for raising your hand when support measures appear.
Read the five roadmap pages for the items that touch you. Every item follows the same structure: current state and targets, the winning approach and investment specifics, then obstacles and the policy package. Five pages per item. The policy package section is effectively advance notice of the support measures to come, available before the budget request stage.
Sort out export control first. This is my opinion rather than something the document says. With dual-use technology and critical materials at the centre of the funding, companies moving into these areas will not avoid classification work. Starting a compliance programme after the money arrives means deadlines and audits landing together. Build the programme first.
Hunting for your own company's name in those 399 pages is more useful than being impressed by the size of the number. If the name is not there, that is a strategic data point too.
In summary
- Cabinet decision of 21 July 2026, comprising a 75-page strategy and a 324-page investment roadmap
- 17 strategic sectors and 62 major products and technologies, with cumulative public and private domestic investment through fiscal 2040 estimated at over ¥370 trillion
- The ¥370 trillion is public and private combined, not government spending
- Ripple effects overlap between items and the document explicitly rules out aggregating them
- Naval vessels and LNG carriers have no calculated figures; two items are restatements
- ¥230 trillion is the capital investment projection, ¥250 trillion the public-private target
- The new investment framework removes request ceilings and secures multi-year funding via a special account and bridge bonds
- "Risk management investment" is economic security in different words, with another country's export controls written in as a premise
Whether this strategy succeeds is not something I can judge yet. But a government naming 62 specific areas where it intends to compete, and publishing the reasoning, gives companies material they can actually use. Criticism can come later. Start by finding out whether your name is in there.
If you want help working out how your products are treated under export control, or where to begin building a compliance programme, talk to us directly.
Footnotes
-
Cabinet Secretariat, "Japan Growth Strategy" (Cabinet decision, 21 July 2026) https://www.cas.go.jp/jp/seisaku/nipponseichosenryaku/pdf/jgs2026.pdf ↩
-
Cabinet Secretariat, "Public-Private Investment Roadmap for Major Products and Technologies in the 17 Strategic Sectors" (Japan Growth Strategy Headquarters, 21 July 2026) https://www.cas.go.jp/jp/seisaku/nipponseichosenryaku/pdf/rm2026.pdf ↩
