On June 24, 2026, UNECE adopted the first global technical regulation for automated driving systems, GTR No. 26, together with UN Regulation No. 185, co-sponsored by Canada, China, the European Commission, Japan, the UK, and the US. AxiSentinel evaluates autonomous vehicle software and safety-case evidence continuously, against whichever jurisdiction's rules actually apply to a given deployment.
Last verified: 23 October 2026
5 industry-specific mandates and 61 that apply across all industries, worldwide.
Next date for this industry: 1 Nov 2026, Egypt PDPL compliance period ends.
See all 66 mandates →Autonomous vehicles have spent a decade being regulated country by country, sometimes state by state. That era is ending: the world's largest vehicle markets just agreed on one definition of a safe automated driving system.
A safety-management-system audit happens once, at type approval. AxiSentinel is built for what happens in the months and years between audits, while the vehicle is actually on the road.
AxiLayer replaces the annual audit snapshot, not the auditor: the human auditor still signs every finding, but the evidence is Always On or On Demand, current the day it is read.
Watch on YouTube →Download the script (PDF) →A vehicle sold globally now has to satisfy a shared international baseline plus whatever each market layers on top. AxiSentinel tracks both levels.
Formally adopted as UN GTR No. 26, paired with UN Regulation No. 185. The first global technical regulation of its kind, endorsed by Canada, China, the EU, Japan, the UK, and the US. Requires audited safety-management systems, credible test environments, and continuous post-deployment monitoring.
The first dedicated AI-safety framework for road vehicles, addressing how AI components inside a vehicle's software stack should be developed and assessed.
China's first mandatory national standard for Level 3 and Level 4 automated driving systems, GB 44721-2026, aligned with the UN GTR for international harmonization.
In force since 15 May 2026, governing permits for automated passenger services ahead of the full Automated Vehicles Act framework targeted for 2027.
The quality management standard Germany’s automotive OEMs and their AV and component suppliers require across product development, production, and service parts, layered on top of the AI-specific safety and type-approval frameworks above, not a substitute for them.
State-level permitting frameworks specific to autonomous vehicle operation, confirmed in force and distinct from existing state AI-transparency statutes. At the federal level, NHTSA withdrew its AV STEP proposed rule on 26 June 2026, citing expected minimal participation in a voluntary program; the agency is pursuing its automated-vehicle safety goals through existing exemption and crash-reporting programs instead.
General-purpose AI risk-management frameworks that already apply to any AI component inside an autonomous vehicle's software stack.
From a single ADAS feature to a fleet of Level 4 robotaxis, AxiSentinel evaluates the software and safety-case evidence continuously, not just at type approval.
AI now helps engineering teams compare architectures, sensor suites and software options during vehicle development. ISO/PAS 8800 sets expectations for the safety of AI in road vehicles, and every AI-assisted design choice becomes part of the safety argument.
Vehicle AI must keep working safely when someone is actively trying to fool or reach it. AxiLayer AI works hand in hand with your cybersecurity team to test, monitor and evidence that these AI systems stay secure, under your written authorization.
Explore AI Cybersecurity Request an AI Cybersecurity Readiness Review
Status as of 28 October 2026. General information, not legal advice. Certifications, CE marking and regulated assessments named here are issued by the bodies the law designates.
Nothing about AxiSentinel's core architecture changes for autonomous vehicles. What changes is which RegDef packages are switched on and what telemetry the agents capture from the vehicle or fleet.
Agents capture vehicle and fleet telemetry on the cadence you configure, Always On or scheduled, in full rather than sampled, and never limited to a quarterly cycle.
Every telemetry event evaluated against the applicable automated-driving rule set, per AI event.
Every audit record is linked to the one before it in a signed, tamper-evident evidence chain, verifiable from the first event.
A compliance state change on one vehicle or an entire fleet propagates network-wide as it happens.
Nothing becomes a compliance finding until a qualified AxiLayer AI auditor reviews and signs it.
AXI-Node agents deploy on vehicle compute, mobile networks, and telecom network functions, with the .axibatch format available for isolated environments.
Tracks human involvement in remote supervision and teleoperation decisions and feeds it into AxiSentinel's oversight-gap scoring model.
The same agents generating compliance evidence watch for spoofed sensor inputs, adversarial perception attacks, and poisoned model updates before a re-certified release reaches production.
New vehicle, jurisdictional, or GTR-aligned rules are added by encoding new RegDef packages. Deployed agents are never rebuilt.
Scoped to your organization during onboarding, not hard-coded into the platform.
A threshold breach becomes a flagged, timestamped, evidence-linked Provisional Alert, reviewed by a qualified AxiLayer AI auditor before anything counts as a finding.
The same architecture monitoring a robotaxi fleet monitors a trading desk's model or a hospital's diagnostic AI. What changes is which RegDef packages are switched on.
The central rule is UN GTR No. 26 on automated driving systems, adopted on 24 Jun 2026 with UN Regulation No. 185 and co-sponsored by Canada, China, the EU, Japan, the UK and the US. The UK permit regime for automated passenger services has applied since 15 May 2026. China's mandatory L3/L4 standard, GB 44721-2026, takes effect on 1 Jul 2027, and ISO/PAS 8800 covers AI safety in road vehicles.
China's GB 44721-2026 takes effect on 1 Jul 2027. Approved on 30 Jul 2026 and published by MIIT on 4 Aug 2026, it is China's first mandatory national standard for Level 3 and Level 4 automated driving systems and is aligned with the UN GTR. New models must comply immediately when the standard is implemented, so type-approval evidence should be ready before that date.
The UN GTR expects continuous post-deployment monitoring and real-world reporting, not point-in-time testing. Manufacturers need an audited safety-management system covering the full life of the automated driving system and evidence that performance matches or exceeds a competent human driver. For robotaxi fleets that includes perception-drift monitoring, disengagement and incident correlation, and logs of the human judgment behind remote assistance.
An AI audit for autonomous vehicles is a third-party assessment of perception, planning and control software and the safety-case evidence behind it, checked against the UN GTR, ISO/PAS 8800 and the rules of each market where the vehicle operates. It continues between type-approval audits, while the vehicle is on the road. A qualified AxiLayer AI auditor reviews and signs every finding.
AxiLayer AI evaluates autonomous vehicle software and safety-case evidence continuously with AxiSentinel, against the rules that apply to each deployment, from the UN GTR baseline to China, UK, Texas and California requirements. Evidence is Always On or On Demand, current the day it is read. A qualified AxiLayer AI auditor reviews and signs every finding. Request a briefing to discuss your vehicle program or fleet.
Last verified: 23 October 2026. See every mandate for this industry on the AI Audit Mandates Map.
Whether it's a single ADAS feature or a fleet of Level 4 robotaxis, AxiSentinel evaluates it the same way it evaluates any AI system: on the cadence you configure, Always On or scheduled, with full evidence, and with a human signature before anything counts as a finding.
