Tesla FSD Is SAE Level 2 — So Why Can’t You Use It in Japan? (July 2026 Update)

Tesla FSD Is SAE Level 2 — So Why Can’t Japan Approve It?

On March 5, 2026, Tesla held an FSD (Full Self-Driving) Supervised test ride event in Shinjuku, Tokyo. Multiple outlets including NTV News, Nikkei, and WIRED Japan attended, and footage of FSD operating on public roads spread widely on social media.

(Source: Soluton TV, March 8, 2026)

Play

(Added March 13, 2026: video by Gomi-san)

Play

Tesla Japan president Hashimoto stated in a Nikkei interview that the company is targeting 2026 implementation and is doing everything possible. Excitement spread on social media: “FSD is finally coming to Japan!”

But here’s an obvious question.

FSD Supervised is SAE Level 2 — the driver must always keep hands on the wheel and bear full responsibility. If an accident happens, it’s the driver’s fault. So why hasn’t Japan’s MLIT simply approved it?

Digging into this reveals a tangled structure where SAE levels and international safety standards operate on completely different axes. This article untangles that structure and assesses how realistic a 2026 Japan launch actually is.

First, let’s establish the baseline. SAE J3016 classifies automated driving into six levels (0-5) based on who bears responsibility for driving.

SAE Level	Who Monitors	Responsibility	Examples
Level 0-1	Driver	Driver	ACC, Lane Keeping Assist
Level 2	Driver	Driver	Tesla FSD Supervised
Level 3	System (limited conditions)	System (conditionally)	Honda SENSING Elite
Level 4-5	System	System	Waymo robotaxis

Tesla consistently classifies FSD Supervised as Level 2. The driver must keep hands on the wheel, eyes on the road, and be ready to intervene at any moment. If an accident occurs, the driver bears responsibility.

By this logic, “Level 2 = driver responsible = same as any other car, just approve it.” In the US, FSD Supervised is sold without NHTSA type-certification via Tesla’s own self-certification. South Korea and Australia have already launched or plan to launch FSD. And in April 2026, the Netherlands RDW formally approved FSD as an EU type-certified system.

But Japan’s regulations don’t operate on SAE levels. That’s the key.

Note: On April 10, 2026, the Netherlands RDW (vehicle certification authority) approved FSD Supervised using the EU type-approval framework (Article 39 exception provision). This is a Netherlands-only approval; EU-wide expansion requires a separate EC application and member state majority vote. The RDW’s official statement clarifies that the EU version of FSD runs entirely different software from the US version, specifically developed for European traffic rules. Since Japan’s regulatory framework is UN-R79-based (distinct from the EU framework), the Dutch approval does not directly apply to Japan — but the fact that “a European authority has approved FSD” could influence Japan’s regulatory discussions.

Japan’s Safety Standard UN-R79: The Structural Barrier Blocking FSD

Japan’s vehicle safety standards adopt UN Regulations. The relevant rule for automated steering is UN-R79 “Steering Equipment,” which classifies automated steering as ACSF (Automatically Commanded Steering Function):

Category	Function	Scope	Status
A	Automated parking	Below 10 km/h	Enacted
B1	Hands-on lane keeping (LKAS)	—	Enacted
B2	Hands-off lane keeping	Highways only	Defined in UN-R171
C	Driver-initiated lane change	Highways only	Enacted
D	System-suggested, driver-confirmed lane change	Highways only	Enacted
E	Continuous automated steering (lane keeping + lane change)	Highways only	Defined in UN-R171

Here’s the critical problem.

FSD Supervised performs continuous automated steering on both public roads and highways — lane keeping and lane changes decided autonomously. Functionally, it is closest to Category E. However, Categories B2 through E are all defined for highways only.

Continuous automated steering on public roads is not covered by any existing category.

The issue is not “Can Level 2 be approved?” but rather “It is Level 2, but there is no applicable category in the safety standard.” Even if Japan’s MLIT wanted to approve it, there is no standard to certify compliance against.

3 Reasons “Level 2 = No Review Needed” Doesn’t Apply to Tesla FSD

You might think: “If there’s no category, grant it a blanket pass as Level 2.” Understandable. But from MLIT’s perspective, there are serious structural risks.

① The precedent risk: “Level 2 = automatic pass”

If FSD Supervised were approved without review simply because it is “Level 2,” any manufacturer could claim “the driver is supervising, so it’s Level 2” to bypass review for any automated steering system. This is an unacceptable moral hazard for a regulatory body.

② OTA updates that continuously change system behavior

Traditional type-certification assumes a fixed spec at the time of certification. FSD Supervised is neural-network-based, and OTA updates continuously alter its behavior. How do you certify a system that could be fundamentally different the following month? UN-R156 provides a framework for software updates, but nothing as fundamental as FSD’s behavioral changes was anticipated.

③ SAE / UN-R79 / UN-R171 / UN-R157 operate on completely different axes

Framework	Classification Axis	FSD Supervised’s Position
SAE J3016	“Who bears responsibility”	Level 2 (driver responsible)
UN-R79 (ACSF)	“What the system does”	Public road continuous steering → no applicable category
UN-R171 (DCAS)	Specifics of UN-R79 B2/E	Highways only → public roads excluded
UN-R157 (ALKS)	“Level 3 or above?”	Level 2 FSD is out of scope

This is the most fundamental point. Four frameworks, four different axes — FSD Supervised does not fit cleanly into any of them. SAE Level 2 (driver responsibility) is not a sufficient condition for Japanese safety standard approval. Note: DCAS Phase 3 is relaxing lane change restrictions, but a separate framework is needed for FSD Supervised’s full capabilities — more on this below.

Why This Gets Confused With “Level 4 Discussions”

The discussion so far can be summarized in the table below.

Regulation	Relationship to FSD	Issue
SAE J3016	Level 2 ✓	SAE level is not Japan’s approval requirement
UN-R79 (ACSF Cat. A〜E)	Public roads → no category ✗	Cat. B2-E defined for highways only
UN-R171 (DCAS)	Highways only → public roads excluded ✗	Phase 3 revision expanding scope
UN-R157 (ALKS)	Level 3+ only → out of scope ✗	Level 2 is out of scope
New ADS Regulation (UN-R number TBD)	Covered via Safety Case △	Targeting formal adoption June 2026

Following news about FSD Supervised’s Japan launch, you often encounter Level 4-adjacent topics — “EU autonomous driving regulations,” “new UN rules,” “UNECE.” Why, for a Level 2 system?

As Not a Tesla App reported in February 2026, UNECE’s upcoming revision could be the approval pathway for FSD Supervised. There are two routes. The primary one is thenew ADS (Automated Driving System) Regulation — primarily targeting Level 3+ systems, but the Safety Case approach makes it a comprehensive framework applicable regardless of SAE level. FSD Supervised’s full capabilities — including turns and intersection handling on public roads — can only be covered through this route. The secondary route isDCAS (UN-R171) Phase 3 revision. At the January 2026 GRVA meeting, system-initiated maneuvers (SIM) on public roads were adopted. However, DCAS “Manoeuvre” primarily covers lane changes — signal-compliant turns and intersection navigation are out of scope. This route is valid for highway use only; full FSD approval requires the new ADS regulation.

Key features of the new ADS regulation:

• SAE-level agnostic“Safety Case” approachadopted
• Approval granted when a manufacturer demonstrates safety superior to human drivers (data-driven)
• US NHTSA, China, and Japan have all expressed intent to align
• Currently under deliberation, targeting formal adoption in June 2026

In short: FSD Supervised has no place in existing UN-R79 categories, so the only path is a comprehensive new regulation that transcends SAE levels. That is why it converges with Level 4 discussions.

WIRED Japan‘s test ride report noted “the key lies in EU developments” and “a lot could move at once from this month onward” — clearly anticipating the June UNECE adoption.

Will Japan Actually Approve FSD by End of 2026?

So is President Hashimoto’s “2026 implementation” target achievable? Below is a timeline with analysis. Notably, the April 2026 RDW approval of FSD could serve as a reference model for Japan’s regulators — “a European authority actually approved it.” The RDW processes approximately 50,000 type approvals annually as one of Europe’s largest certification bodies, and its safety assessment data could be valuable reference material for Japan’s MLIT.

Optimistic Scenario (approval in late 2026)

1. ADS new regulation formally adopted at the June UNECE session
2. MLIT rapidly incorporates it into domestic standards
3. Tesla submits safety data and obtains type certification
4. FSD delivered via OTA to approximately 40,000 existing Tesla vehicles in Japan

In this scenario, Japanese Tesla owners could use FSD Supervised as early as fall-winter 2026. Tesla held a Shinjuku media test ride and President Hashimoto publicly stated the company is “doing everything possible” — suggesting Tesla may already have confidence that this timeline is achievable.

Cautious Scenario (slips to 2027 or later)

• UNECE new regulation adoption is delayed (common in international negotiations)
• Even if adopted, MLIT takes time to reflect it in domestic standards
• Additional data required for compliance with Japan-specific traffic rules (crosswalk stops, railway crossings, etc.)

Frankly, Japan’s regulators have never been known to take first-mover risk —  their track record speaks for itself. Even after the ADS regulation is adopted, immediate domestic application is far from certain. However, as operational data accumulates from the Netherlands deployment, the track record of “FSD actually running safely in Europe” could serve as supporting evidence for Japanese regulators’ approval decisions.

When Safety Case Unlocks Japan

For Tesla owners in Japan, FSD launch is the most anticipated event. Having a major feature light up via OTA is the kind of moment Tesla owners live for. HW3 vehicles should receive FSD shortly after HW4, with a slight delay.

Having researched this thoroughly, the conclusion is clear: this is not as simple as “it’s Level 2, just approve it.” The structural mismatch between SAE’s “responsibility framework” and Japan’s safety standard “function regulation” cannot be resolved by Tesla’s efforts alone.

Conversely, if the June UNECE new regulation is adopted, this structural mismatch could be resolved all at once. The “Safety Case” approach — approve it if it is safer than a human — is the framework most suited to FSD Supervised. The vast driving data Tesla has accumulated in the US can serve directly as safety evidence.

[June 11, 2026 Update] On the Eve of the UN Vote — Reading the WP.29 Agenda

The optimistic scenario above rested on one condition: “the UNECE meeting in June formally adopts the ADS regulation.” That meeting now has a date. The agenda for the 199th session of UN WP.29 (ECE/TRANS/WP.29/1191) has been published: Geneva, June 23–26, with votes on June 24.

The Premise Has Shifted — Five Countries, While the EU Vote Stalls

A quick premise update first. Starting with the Dutch RDW approval on April 10, mutual recognition has chained through Lithuania (May 20), Estonia (May 29), Denmark (June 9) and Belgium (June 10) — five approved countries in eight weeks. Meanwhile the EU-wide vote in the TCMV has not advanced: no vote on May 5, no vote slot on June 30, with a realistic bloc-wide vote now expected in October or later. The country-by-country details are tracked in our Netherlands approval article. What matters here is a single point: while Europe’s operating record accumulates, the international rulebook itself moves this month.

A Two-Story Voting Machine

A WP.29 session is not one meeting. The World Forum’s plenary hosts the voting committees of separate treaties: agenda item 4 covers the 1958 Agreement (mutual recognition of type approvals — Japan and the EU are parties), item 5 the 1998 Agreement (Global Technical Regulations, which include the US and China), and item 7 the 1997 Agreement (periodic inspections). Then item 11 convenes AC.1, the voting committee of the 1958 Agreement, and item 14 convenes AC.3 for the 1998 Agreement. On June 24, automated-driving standards go to a vote in both worlds — the Japan/EU type-approval world and the US/China technical-regulation world — on the same day.

What’s on the Table on June 24

Picking out the automated-driving items shows the sheer volume:

Item	Content	Why it matters
4.14.2	New UN Regulation on ADS	The international standard for Level 3+ automated driving — one of three new regulations on the agenda
14.1.1	UN GTR on ADS	The same substance voted under the 1998 Agreement, covering the US and China
4.7.16	UN R171 (DCAS) — 02 series	Expansion of the Level 2 regulation directly relevant to FSD Supervised; the Dutch approval combined R171 with an Article 39 exemption
4.7.8	UN R79 — new 05 series	The steering regulation that blocked FSD in the first place, now itself being revised
4.6–4.11	FADS amendment package	Roughly 60 documents adapting existing regulations (braking, steering, speedometers…) for automated vehicles
8.2	Consistency with the 1949/1968 Road Traffic Conventions	The deepest issue: treaties that presuppose a “driver” versus driverless operation

Three details deserve attention. First, the volume of the FADS package: about 60 amendment documents, finalized by a joint committee as recently as June 8–10, that rewrite the existing rulebook for automated vehicles rather than just adding one new regulation. Second, R79 is on the same table. As this article explained, R79’s constraints on continuous automated steering are what blocked FSD; with R171’s 02 series and R79’s 05 series updated on the same day, Level 2 advanced assistance gets a proper regulatory home from two directions. The front-door route for FSD’s full capability remains the ADS regulation, but the R171 expansion reduces dependence on the Article 39 “exception” route. Third, item 8.2 is a quiet bombshell: Japan is a party to the 1949 Geneva Convention, and the “driver” concept in its Road Traffic Act derives from it. However complete the vehicle standards become, the treaty layer remains.

The vote was never a guaranteed outcome. In the GRVA deliberations, China had questioned whether the frequent amendments to R171 were backed by evidence that earlier versions were sufficiently safe, and Norway had entered reservations on the draft. In international standards, “tabled” and “adopted” remained different things.

What It Means for Japan — Three Axes

Japan’s path runs on three axes. First, international standards. If the R171 expansion, the R79 revision and the new ADS regulation pass on June 24, MLIT has the full toolkit to incorporate into Japan’s safety standards. Regulations under the 1958 Agreement typically enter into force about six months after adoption — making early 2027 the institutional milestone. Second, operating record. Real-world data from five European countries is the best possible evidence for a regulator that, as this article noted, does not take world-first risks. Third, Japan-specific localization. This one remains: crosswalks, railway crossings, narrow residential streets. European approvals solve none of Japan’s edge cases — and as item 8.2 shows, beyond vehicle standards lies the separate layer of traffic law.

Hashimoto’s “within 2026” has moved from pure aspiration to conditionally within range, as the first two axes advance this month. But as the TCMV stall shows, institutions move slower than technology. The optimistic “fall–winter 2026” remains ambitious; this article’s position stays with the cautious 2027 reading. We will update after the June 24 vote. (Day-to-day developments are tracked on our sister site tesla.lounges.site — an approval matrix and a UN/EU regulatory meeting calendar, in Japanese.)

[June 24, 2026 Update] The World’s First ADS Rules Were Adopted — But Only for Driverless Systems

They were adopted.

In the June 11 update, this article’s optimistic scenario put “formal adoption of the new ADS regulation at the June UNECE session” as condition number one. The night before the vote, I also cautioned that “being on the agenda and being adopted are different things.” On June 24, that condition was met at WP.29’s 199th session. The ADS regulation itself and a package of roughly 90 accompanying amendments were adopted together. Reservations raised by China and Norway during deliberations did not stop the outcome.

—And yet, opening the adopted text was a surprise. It was not the “shape” I had expected.

Until the Text Was Final, No One Could Read the Scope

To be honest: before adoption, whether this ADS rule would cover Level 2 at all was unreadable until the final text appeared. Draft-stage commentary widely treated the Safety Case approach as a level-agnostic, comprehensive framework — and the June 11 update bet on that possibility, writing that “today, only this route can cover full FSD Supervised.”

On June 24, the ambiguity resolved. The finalized ADS regulation is, by definition, Level 3–5 only. The text defines “Automated Driving System” as systems that perform Levels 3, 4, and 5. An ADS is a system that “sustains the entire Dynamic Driving Task (DDT).” FSD Supervised — where the driver keeps monitoring and remains responsible, and which Tesla itself positions as Level 2 — sits outside that definition. UNECE’s press release, from headline to body, consistently said “fully driverless.”

In other words, the world’s first common automated-driving rule set excluded, at the door, the product Tesla had brought closest to Japanese roads. Having braced for “this is what unlocks Japan FSD,” it was a clean miss.

What Actually Moved — A Two-Tier Structure

There is no need to despair. Once the sequence is clear, the picture sharpens.

June 24 was not one event but two tiers. First, a global common foundation for driverless operation (Level 3+) was laid for the first time. Second, in parallel, a bundle of amendments aligning roughly 90 existing rules with the ADS era was adopted — the unglamorous but essential “plumbing” package this article has described. UN-R79, which has blocked FSD, appears to fall within that amendment set’s reach (individual rule breakdowns will be confirmed against primary documents).

The flagship ADS regulation aims at Unsupervised FSD, Robotaxi, and Cybercab — layers without a human driver. Japan’s home for FSD Supervised (Level 2) is not settled by this main rule. That homework remains. The structure this article has argued all along — “SAE levels and approval frameworks move on different axes” — was, ironically, confirmed by the international text itself.

When Does It Enter Into Force? — The Dates Still Diverge

Entry-into-force timing still has a spread. UNECE’s press release says “about one month.” AFP quoted GRVA secretariat’s Guichard as expecting “January 2027.” International entry into force and each country’s incorporation into domestic law (Japan: safety standards and the Road Traffic Act) are different processes. Japan mainly absorbs 1958 Agreement UN Regulations; after adoption, domestic drafting begins. Given a regulator that does not take “world-first” risk lightly, the rest remains a race against time.

So Where Does FSD Live?

What Moved Was the “Driverless” Axis

To restate: conditions for FSD Supervised in Japan still sit on three axes — SAE level, UN regulations (safety standards), and traffic-rule treaty alignment. What moved sharply on June 24 was the UN-regulation axis — but its tip extended toward driverless (Level 3+). Newly created main rule UN-R185 targets Level 3+; Level 2 FSD Supervised was left outside.

The sequence is interesting. Common sense says “driver assist (Level 2) first, then driverless (Level 4).” International standards went the other way: a common driverless rule (UN-R185) first as the global main body, while advanced Level 2 assist stayed in existing individual regulations. For Japanese Tesla owners, a different door opened first than the one they had been watching.

That other door — the newly adopted framework — does not, on inspection, license optimism for FSD Supervised Japan. The automated-driving main body adopted this time was not an R79 amendment; it was created as new rule UN-R185. Scope is M1–N3 vehicles and Level 3+ ADS (numbering is from the adopted proposal text; final confirmation awaits the session report). R79’s structural hole — no category for continuous automated steering on public roads — was not fixed by rewriting R79; separate R185 fills the driverless side instead. R79 remains. So where is the real home for Level 2 FSD Supervised? It is the other box: DCAS (UN-R171). Europe’s Tesla approvals rest on R171 conformity. Japan’s MLIT also positions DCAS as “equivalent to Level 2 assist, including public-road lane changes and intersection turns.” The public-road core is entering range. The remaining gap is hands-off and system-initiated maneuvers.

In the End, a “Top-Down Decision”

Summary: the driverless (Level 3+) main body is new UN-R185. The Level 2 FSD Supervised box is DCAS (UN-R171). The box exists. The problem is that bleeding-edge behaviors such as hands-off still do not fully fit inside it.

Here Europe and Japan diverge decisively. Europe has Article 39 — a provisional path for behaviors that do not fit the rules, by authority judgment. Tesla’s Dutch approval used R171 as base and filled the overflow with Article 39. Japan has no commercial equivalent of Article 39. Trial-use relaxation exists, but it is a different animal.

Even if Japan incorporates DCAS (R171) 02 series into safety standards, what expands in orderly fashion is highway hands-off and system-initiated maneuvers centered on lane change. The true value of FSD Supervised — public roads, hands-off through signals and intersections — stays out of reach on the regular DCAS path for some time. Without a European-style provisional escape hatch, full public-road FSD will not run in Japan.

How Japan prepares that escape hatch is not a matter of bureaucratic stack-up or pure technology. The decision to step in is Japan’s version of a “top-down call.” That is how I see it. I chased this as a technology story; the end of the trail is institutional design and human decision.

The thesis that “levels and approval frameworks are different axes” is more firmly supported now that the rules and national practice are open. Continuous monitoring continues on the FSD Approval Tracker.

[July 13, 2026 Update] UN-R171’s 02 Series Was Adopted Too

On the June 24 vote, this article reported adoption of the ADS regulation (UN-R185). But the same day, another box was updated: agenda item 4.7.16 — UN-R171 (DCAS) 02 series amendments. The June 11 update listed it on the agenda table and then failed to recover the result. This is the catch-up.

The 02 series is what the main text called “Phase 3 amendments.” Provisions that limited system-initiated maneuvers (SIM) to motorways were deleted, and test items premised on non-motorway operation were added. The draft was adopted at the 24th GRVA session in January and finalized at WP.29 on June 24. International entry into force is reported as 2 October 2026; primary-document confirmation continues.

The prior update’s framing advances one step. I wrote that “the Level 2 box is DCAS (R171); the problem is that hands-off and SIM still do not fully fit.” With 02 series adoption, extending SIM — system-proposed lane changes and the like — onto public roads has entered the regulatory box. Part of what Europe handled via Article 39 overflow now sits inside the formal frame.

The Japan implication is timeline. Domestic incorporation of R171 has precedent: 00 series adopted WP.29 March 2024 → safety-standard amendment September 2024; hands-off amendments adopted March 2025 → domestic take-up timed to September 2025 entry into force. A roughly six-month cycle twice. At the same pace, 02 series domestic reflection points to late 2026–early 2027.

The spine of the conclusion does not change. What 02 series expanded is maneuvers centered on lane change; signal-following turns and intersection handling — the core of FSD Supervised — remain out of scope. What changed is that the floor of a “highway-only compromise product” rose to “public-road lane changes included.” Scenario ① in the FAQ became only slightly more productizable.

Conclusion

The key takeaways from this analysis:

1. FSD Supervised is Level 2 (driver responsible), but Japan’s safety standard (UN-R79) has no category for continuous automated steering on public roads.This is why “Level 2 = approvable” does not hold.
2. There are two possible routes, but the primary one is the UNECE ADS new regulation targeting June 2026 adoption (UN-R number TBD).The SAE-agnostic Safety Case approach makes it the only framework capable of approving FSD Supervised’s full capabilities including turns on public roads. DCAS (UN-R171) Phase 3 is primarily a lane change relaxation — a partial approval route limited to highway use.
3. 2026 implementation is not impossible.The UNECE regulation adoption and MLIT’s response speed are the decisive factors.

We will continue to watch developments around the Netherlands RDW approval, whether it expands EU-wide, and the June UNECE formal adoption. Until then, keep enjoying Autopilot — and keep your hands on the wheel.

---

References
Nikkei: “Tesla Targets 2026 Implementation of AI Autonomous Driving in Japan”
WIRED Japan: “Tesla’s Autonomous Driving Targeting 2026 Implementation in Japan”
Not a Tesla App「New UN Autonomous Driving Regulation Could Unlock Tesla FSD in Europe and Asia」
Cabinet Office SIP Automated Driving: Status of International Standards for Automated Steering
MLIT: Review Status of UN-R79 Revision
Tesukasu: “MLIT Approves OTA Activation of Automated Driving Functions”
Drive Tesla Canada「Tesla targets FSD launch in Japan by late 2026」
RDW: Official Statement on European Type Approval for Tesla (Netherlands Vehicle Certification Authority)

Timeline Toward FSD Supervised Approval in Japan (2017-)

Date	Event	Notes
2017-2021	UN-R79 ACSF Cat. A-C enacted	Adopted into Japan’s safety standards. Defines hands-off lane keeping (B1), automated lane change (C), etc.
2021.03	UN-R157 (ALKS) enters into force	International standard for Level 3 highway driving. Honda SENSING Elite becomes first Japan-certified Level 3.
2022.07	Japan starts applying UN-R155/R156	Cybersecurity standards for OTA updates. Applied to new models first.
2025.08	Tesla FSD internal testing begins in Japan	Model 3 driving in Tokyo. Tesla employee as safety driver.
2025.10	MLIT: Approves adding driver assistance functions via OTA	FSD delivery to sold vehicles becomes technically possible. Separate from FSD approval itself.
2026.01	UNECE ADS new regulation draft adopted (WP.29)	Comprehensive Safety Case approach rule. US NHTSA, China, Japan express intent to follow.
2026.01	DCAS (UN-R171) Phase 3 SIM restriction relaxation adopted (GRVA)	Relaxes restrictions on system-initiated lane changes. Directly relevant to FSD Supervised.
2026.03	FSD Supervised Shinjuku Media Test Ride	Demo with Model Y Juniper. President Hashimoto: “Targeting 2026 implementation.”
2026.04	Netherlands RDW approves FSD Supervised as EU type-certified system	Article 39 exception provision. EU version runs different software from US version. EU-wide expansion requires EC application.
2026.05?	UN-R155/R156 full application to continuing production vehicles	Applied to all vehicles including non-OTA. Regulatory foundation complete.
2026.06.24	WP.29 199th session — ADS regulation (UN-R185) and R171 02 Series adopted. FSD Supervised (Level 2) remains outside UN-R185’s scope	Takes effect about six months after adoption, per standard 1958 Agreement procedure. FSD Supervised approval pathway opens.
Late 2026?	Japan FSD Supervised approval?	Contingent on DCAS Phase 3 domestic adoption or ADS new regulation domestic adoption.

---

Frequently Asked Questions

Related Articles

• The Truth Behind Honda’s EV Exit: ¥690B Loss and the Strategic Failure of Honda 0 (Japanese)
• The Loyalty Test Called Clean Energy Subsidy — The Facade, the Real Agenda, and the Exit Strategy of Japan’s CEV Subsidy