2026-06-18 — views
Physical AI Regulatory Landscape — AV Permit Map, NHTSA Framework, and EU Type-Approval Roadmap for Waymo and Tesla FSD
Waymo holds full driverless permits in CA, AZ, TX; Tesla uses Texas self-cert only. EU Level 3-plus type-approval is the largest remaining regulatory unlock.
Article 141 in the Physical AI Benchmark Series — Physical AI Regulatory Landscape: State-by-State AV Permit Map, NHTSA Federal Framework, and EU Type-Approval Roadmap for Tesla FSD and Waymo
Permits are the operational chokepoint in the Physical AI race. The best autonomous vehicle technology in the world cannot generate a single fare, mile of driverless commercial operation, or unit of subscription revenue without regulatory authorization. This article maps the full regulatory landscape across three jurisdictions — US state permitting, NHTSA federal framework, and EU type-approval — and scores Waymo and Tesla FSD against every dimension.
All figures labeled “(est.)” are derived from public disclosures, regulatory filings, research publications, and industry analyst estimates rather than independently verified primary data. This article does not constitute investment advice.
Section 1 — US State-by-State AV Permit Map
The United States has no uniform federal AV deployment framework as of mid-2026. Operational rules are set state by state, creating a patchwork of requirements that ranges from California’s multi-stage public-hearing permit process to Texas’s operator self-certification model. This patchwork is the primary reason Waymo’s city-by-city expansion pace is approximately one new city per year: each new state requires a separate regulatory engagement strategy.
| State | Driverless commercial rides allowed | Key requirement | Waymo status | Tesla FSD status | Notes |
|---|---|---|---|---|---|
| California | Yes (CPUC permit + DMV driverless deployment permit) | CPUC driverless deployment permit; NHTSA FMVSS exemption not required for modified vehicles | Commercial driverless (SF, LA, San Jose areas) | No driverless permit; supervised only | California has the most rigorous AV permitting in the US; Waymo holds full commercial driverless permit; Tesla has not applied for CA driverless permit |
| Arizona | Yes (ADOT permit framework; self-certified) | Annual reporting; no fleet size cap; no CPUC-style public hearing | Commercial driverless (Phoenix/Chandler/Tempe/Scottsdale) | Supervised FSD; no driverless operation | AZ most permissive early-mover state; Waymo’s largest fleet by vehicle count |
| Texas | Yes (HB 1791, 2017; self-certification model) | Operator self-certifies compliance with Texas Transportation Code; no state agency approval required | Commercial driverless (Austin) | Tesla Austin robotaxi launched under TX self-cert | Texas is the most permissive large-state AV framework; Tesla’s robotaxi launch leveraged this specifically |
| Nevada | Yes (DMV AV testing + commercial permit) | Annual reporting; NV DMV approval required | Testing only (no commercial rides as of mid-2026) | Supervised testing; no commercial driverless | Nevada was first US state to pass AV law (2011) but commercial deployment lagged behind CA/AZ/TX |
| Florida | Yes (HB 7027; self-certification) | No state agency approval; operator assumes liability | Limited testing by multiple companies | Supervised FSD; no commercial driverless | FL framework similar to TX; warmer weather and light regulation attracting testing |
| Michigan | Yes (PA 231, 2016; self-certification + reporting) | Annual report to MI Dept of State Police | Testing; no large-scale commercial driverless | Supervised FSD; no commercial driverless | Motor City regulatory sandbox; focus on testing not commercial ops |
| Washington D.C. | Yes (DC AV Act; permit framework) | DC permit + reporting | No commercial presence | Supervised FSD; no commercial driverless | Small jurisdiction; high regulatory visibility due to federal proximity |
| All other US states | Generally requires legislation or executive order; many have neither | Varies widely; 29 states have some AV legislation (est.) | Not present commercially | FSD supervised where not prohibited | AV regulatory patchwork creates city-by-city expansion complexity |
California: The Hardest Permit and Why It Matters Most
California’s multi-agency AV permit structure — requiring both a California DMV driverless deployment permit and a California Public Utilities Commission (CPUC) transportation network company permit for paid rides — is deliberately rigorous. The CPUC process includes public comment periods, safety hearings, and ongoing incident reporting requirements. Waymo spent multiple years navigating this process before receiving its full commercial driverless paid-rides permit. The significance: a California driverless permit is the most credible safety credential in the US market. Tesla has not applied for a California driverless permit as of mid-2026.
Texas: The Self-Certification Model and Why Tesla Used It
Texas HB 1791 (2017) created the most permissive large-state AV legal framework in the US. Under Texas law, an AV operator self-certifies compliance with the Texas Transportation Code, files no pre-deployment application with any state agency, and faces no fleet size cap. This is why Tesla chose Austin as the launch site for its robotaxi service: the Texas self-certification model allowed Tesla to begin commercial driverless rides with a simple declaration of compliance rather than a multi-year regulatory approval process. The downside is credentialing: the Texas self-cert carries no independent regulatory safety verification, which matters for expansion to stricter states.
Section 2 — NHTSA Federal Framework and FMVSS Exemptions
The Federal Motor Vehicle Safety Standards (FMVSS) set by NHTSA cover vehicle hardware safety — things like crash structures, lighting, braking, and occupant protection. NHTSA does not set operational rules for where AVs can drive; that remains with states. But NHTSA’s authority over vehicle design creates a critical federal bottleneck for any AV company building purpose-built driverless vehicles without traditional steering wheels or pedals.
| Regulatory element | Current status | Impact on Waymo | Impact on Tesla | Notes |
|---|---|---|---|---|
| NHTSA AV jurisdiction | NHTSA sets FMVSS; states set operational rules | Waymo Gen 6 purpose-built vehicle requires FMVSS exemptions (no steering wheel/pedals) | Tesla vehicles comply with all FMVSS as standard consumer vehicles | Federal/state split: NHTSA covers vehicle safety; states cover deployment rules |
| FMVSS exemption (steering wheel/pedals) | NHTSA can grant exemptions for up to 2,500 vehicles/year per manufacturer for safety research; Volume Production Exemption in SAFE 2.0 would allow higher volumes | Waymo has received NHTSA exemptions for pedal-free Jaguar I-PACE fleet; Gen 6 will require new exemption | N/A — Tesla Cybercab (pedal-free design) will need FMVSS exemption similar to Waymo | FMVSS exemption is the federal bottleneck for pedal-free AV volume production |
| AV START Act (proposed) | Multiple AV bills introduced in Congress since 2017; none passed into law as of mid-2026 | Would create federal deployment framework preempting state patchwork; Waymo lobbied for passage | Tesla and Waymo both support federal framework to replace state-by-state approach | Congressional gridlock has left state patchwork as the de facto framework |
| NHTSA AV investigations | NHTSA has authority to investigate AV incidents; opened investigations into Tesla FSD (2021–2022; closed without recall); investigating Waymo (closed without recall 2024) | Waymo investigation closed without recall; reinforces safety record | Multiple FSD investigations closed; NHTSA monitoring but not blocking deployment | Investigations are compliance checks, not deployment blocks |
| Crash reporting requirement (Standing General Order) | NHTSA requires AV companies to report crashes within 1 day (Level 2+) and 10 days (Level 3+) | Waymo reports all commercial crashes under SGO | Tesla reports FSD and Autopilot crashes; largest reporter by absolute volume due to fleet size | High absolute crash counts for Tesla reflect fleet size, not rate; NHTSA evaluates rate, not count |
| SAFE 2.0 Act (signed 2022) | Infrastructure Investment and Jobs Act; pilot program for AV deployment; no binding operational rules | Waymo eligible for federal AV pilot grants | Tesla eligible for federal AV pilot grants | Mostly funding; does not preempt state operational rules |
The FMVSS Exemption Bottleneck
The 2,500-vehicle-per-year FMVSS exemption cap is the federal ceiling on pedal-free AV production until Congress passes expanded exemption authority. Waymo navigated this for its Jaguar I-PACE fleet and will need to navigate it again for the Gen 6 vehicle. Tesla’s Cybercab, which is designed without traditional pedals and steering wheel for its robotaxi configuration, will face the same bottleneck. The Volume Production Exemption proposed in SAFE 2.0 and subsequent legislative proposals would raise this cap significantly — but no such expansion has passed as of mid-2026.
Section 3 — EU Type-Approval Roadmap for AV
The EU type-approval system is the single largest regulatory unlock remaining in global AV deployment. EU type-approval — governed by the UNECE Working Party 29 (WP.29) framework — is binding across EU member states and recognized in 60-plus countries worldwide. A single EU type-approval would unlock commercial AV deployment across a 450-million-person market.
| Regulatory element | Current status | Impact on Tesla FSD | Impact on Waymo EU expansion | Notes |
|---|---|---|---|---|
| UNECE WP.29 framework | UN Economic Commission for Europe Working Party 29 sets vehicle type-approval standards for EU + 60+ countries; binding in EU | FSD requires UNECE WP.29 type-approval for Level 3+ operation; no approval granted as of mid-2026 | Waymo has no EU commercial presence; same regulatory requirement would apply | WP.29 covers 55% of global vehicle market; EU type-approval is the largest single regulatory unlock |
| UNECE Regulation 157 (ALKS) | Automated Lane Keeping System — first WP.29 AV regulation; allows Level 3 up to 60 km/h on highways; adopted 2021 | Tesla FSD could qualify for ALKS at Level 2+ but Level 3 approval under R157 is pending; limited to 60 km/h highway | N/A (Waymo is urban, not highway) | R157 is the entering wedge for EU AV approval; Tesla’s FSD scope is broader |
| EU AI Act (2024) | High-risk AI system classification includes AV; requires conformity assessment, transparency obligations, data governance | Tesla FSD classified as high-risk AI; conformity assessment required before deployment at scale | Same classification applies to Waymo | EU AI Act adds compliance layer on top of type-approval; independent process |
| EU General Safety Regulation (GSR 2) | Mandatory ADAS features for new EU vehicles from 2024; no Level 3+ mandate yet | Tesla vehicles sold in EU must comply with GSR 2 (intelligent speed assistance, emergency braking); FSD as Level 3+ is a separate regulatory question | N/A | GSR 2 compliance (Level 2 ADAS) is table stakes; does not unlock Level 3+ deployment |
| EU member state pilot programs | Germany (AVG 2021), UK (post-Brexit AEVA/AV Bill), France (Loi LOM) allow limited AV testing/trials | Tesla testing FSD in select EU markets; no commercial driverless deployment approved | No EU commercial presence | Member state pilots are testing frameworks, not commercial deployment permits |
| Timeline estimate for EU FSD approval (est.) | No formal timeline; regulatory experts estimate 2027–2030 (est.) for Level 3+ commercial approval in EU under a unified framework | EU approval would unlock 450M population market; estimated 5–10x TAM expansion for FSD subscription | Same unlock for Waymo EU expansion, but mapping requirement per city adds 6–18 months per city | The EU unlock is the largest single regulatory event remaining in global AV deployment |
Why the EU Approval Timeline Is the Defining Regulatory Event
The EU type-approval process for Level 3+ autonomous vehicles requires a combination of UNECE R157 compliance, EU AI Act conformity assessment, and in some cases member-state pilot certification. No AV company has received full EU type-approval for commercial driverless operation as of mid-2026. The regulatory expert consensus estimate of 2027–2030 (est.) reflects the multi-agency, multi-year nature of the EU approval process. For Tesla, EU approval would unlock FSD subscription revenue across a market comparable in size to the entire US. For Waymo, EU approval would require the additional step of HD mapping every target city — adding 6–18 months per city beyond the type-approval date.
Section 4 — Waymo vs Tesla Regulatory Position Comparison
| Dimension | Waymo | Tesla | Edge | Notes |
|---|---|---|---|---|
| US driverless permits | Full driverless: CA + AZ + TX (Austin) | TX self-cert only (Austin robotaxi) | Waymo decisive | 3-state driverless vs 1-state self-cert |
| FMVSS exemption | Granted for Jaguar I-PACE fleet; Gen 6 pending | N/A (current vehicles compliant; Cybercab will need exemption) | Waymo (navigated successfully); Tesla (ahead of curve with Cybercab planning) | Both face federal exemption process |
| EU regulatory position | No EU commercial presence; same WP.29 requirements apply | FSD in EU testing; EU type-approval pending (~2027–2030 est.) | Even | Neither has EU commercial driverless approval |
| State expansion speed | ~1 new state per 2–3 years (est.) — requires mapping + permit application per state | TX self-cert model allows rapid expansion if other states adopt TX-style framework | Tesla (potential speed advantage if TX-style laws spread) | AV STAR Act passage would benefit Tesla disproportionately |
| Regulatory strategy | Proactive engagement: Waymo works with each regulator; builds trust over years | Self-certification in permissive states; avoids proactive engagement in stricter states | Waymo (trust-building); Tesla (speed in permissive states) | Different strategies for different risk tolerances |
| Incident track record with regulators | NHTSA investigation closed without recall; strong relationship with CA DMV/CPUC | Multiple NHTSA investigations closed; mixed relationship with CA regulators | Waymo (cleaner regulator relationship) | Trust is a non-financial permit input |
| Federal lobbying position | Supports federal AV framework; part of AV coalitions | Supports federal framework; less active in traditional lobbying coalitions | Waymo (more engaged) | Federal framework passage would benefit both |
Section 5 — Regulatory Benchmark Scorecard and Ramp Implications
| Dimension | Waymo | Tesla | Verdict |
|---|---|---|---|
| Current US driverless permits | 3 states, full commercial | 1 state, self-cert | Waymo decisive advantage today |
| Permit acquisition pace | Proven: CA (hardest), AZ, TX all achieved | TX first; pace to next state undefined | Waymo leads; Tesla’s next permit is the key watch signal |
| Federal framework risk | Low: would standardize what Waymo already has | Medium: could accelerate or complicate depending on final rules | AV STAR Act passage = Tesla benefit |
| FMVSS exemption (pedal-free vehicles) | Navigated successfully for I-PACE; Gen 6 in progress | Cybercab will need exemption; not yet filed (est.) | Waymo ahead on execution; both face same process |
| EU approval timeline | 2027–2030 (est.) — same as Tesla | 2027–2030 (est.) — largest single unlock remaining | Even; EU is the defining regulatory event of the next 3 years |
| Overall regulatory verdict | Waymo has the deepest US regulatory moat — built over 10+ years of structured regulator engagement, safety data sharing, and permit-by-permit trust accumulation. This moat is real and takes years to replicate. | Tesla’s regulatory position is strong in permissive states and improving nationally, but the gap to Waymo’s multi-state driverless permit portfolio is 12–24 months (est.) at minimum. The EU unlock is the largest single addressable regulatory event for both — whoever gets EU type-approval first gains access to 450M+ population. | Waymo wins regulatory depth; Tesla wins regulatory speed-in-permissive-states |
The Two Scenarios That Change Everything
Scenario A — Federal AV framework passes: If the AV STAR Act or a successor bill passes Congress and creates a federal deployment framework, Tesla’s self-certification model in permissive states becomes the national default. Tesla’s regulatory gap to Waymo narrows from years to months. This is the single legislative event most likely to accelerate Tesla’s driverless ramp.
Scenario B — EU type-approval granted to either company: The first company to receive EU type-approval for Level 3+ commercial operation gains access to 450 million people, a market comparable in size to the entire US. For Waymo, EU approval plus per-city HD mapping means 6–18 months per city to open service. For Tesla, EU approval means FSD subscription revenue begins immediately across all markets where Tesla vehicles already operate — no per-city infrastructure required. Tesla’s mapless architecture gives it a structural advantage in EU ramp speed once regulatory approval is achieved.
Note: All figures labeled “(est.)” are derived from public disclosures, regulatory filings, research publications, analyst estimates, and industry reports as of mid-2026. This article does not constitute investment advice.
Sources
- California CPUC driverless deployment permits — California Public Utilities Commission ↗
- NHTSA Automated Vehicle Policy — US DOT ↗
- UNECE WP.29 AV regulations — UN Economic Commission for Europe ↗
- EU AI Act — European Parliament ↗
- Texas HB 1791 AV framework — Texas Legislature ↗