National and State-Level Approaches to Automated Vehicle Certification: A Comparative Framework for State Operational Governance

Sk Nahia Ahsan, Md Abdullah Al Hasan, Danjue Chen, George ListarXiv 2026 · arXiv

This paper provides a much-needed map of the fragmented regulatory landscape for automated vehicles (AVs) in the United States. By analyzing how different states authorize SAE Level 4 and 5 systems, the authors propose a three-tier typology that categorizes state governance from "minimalist" to "permit-based," offering a strategic framework for policymakers to balance innovation with public safety.

TL;DR

The U.S. lacks a unified federal framework for certifying fully automated vehicles, leaving states to create a "patchwork" of operational rules. This paper develops a three-model typology to categorize these state-level approaches and compares them to centralized international models, providing a roadmap for states to evolve their oversight as driverless technology matures.

What Problem Does This Paper Address?

The emergence of Automated Driving Systems (ADS) at SAE Levels 4 and 5—where the computer, not a human, is in control—breaks traditional safety frameworks. Old regulations were designed for human-operated, mechanically oriented vehicles. Today’s AI-driven systems require continuous oversight of software updates, cybersecurity, and "edge-case" performance that static, one-time hardware tests cannot capture.

In the U.S., the National Highway Traffic Safety Administration (NHTSA) oversees vehicle design via self-certification, but the authority to grant operational permissions (testing, deployment, and insurance) rests with individual states. This has led to significant regulatory fragmentation. Industry leaders cite the lack of a unified federal framework as a major hurdle to safe, large-scale deployment.

Methods

The authors conducted a systematic review of publicly available state statutes, administrative regulations, and agency guidance documents. They selected ten U.S. states (CA, DC, PA, TX, WA, AZ, GA, NC, WY, and MI) to capture the full spectrum of regulatory approaches.

To structure the comparison, they developed an analytical framework based on four dimensions:

  1. Scope and Operational Conditions: What models are used (e.g., permits vs. self-certification)?
  2. Governance and Oversight: Which entities are responsible and what evidence is required?
  3. Federal Interaction: How do state rules mesh with NHTSA guidelines?
  4. Implementation Outcomes: Which companies are actually operating under these rules?

The study also reviewed international practices in the European Union, China, Japan, and Australia to provide a global context.

Datasets

As a policy and governance study, this paper does not use traditional machine learning training sets. Instead, its "dataset" comprises:

  • State-level legal documents: Statutes and administrative rules from the 10 selected jurisdictions.
  • Agency Guidance: Documents from state DOTs and DMVs.
  • Expert Interviews: Insights from NCDOT-sponsored research conducted between January and February 2026.
  • International Standards: UNECE regulations (R155, R156, R157) and ISO standards (26262, 21448).

Key Insights and Ideas

  • The Three-Model Typology: The authors successfully clustered state practices into three distinct tiers based on regulatory workload and institutional complexity:
  • Tier 1 (Baseline Legal Compliance): Minimalist approach where AVs simply follow existing traffic and insurance laws (e.g., Georgia, North Carolina).
  • Tier 2 (Administrative Oversight): Requires AV-specific safety management plans and post-deployment crash reporting, but no formal "permit" to start (e.g., Arizona, Washington).
  • Tier 3 (Permit-Based Authorization): The most rigorous level, requiring formal agency review and an explicit permit before any testing or deployment (e.g., California, Pennsylvania).
  • Elaboration vs. Stringency: The paper distinguishes between how detailed a program is (elaboration) and how restrictive it is (stringency). For example, California is highly elaborate and prescriptive, while the District of Columbia is highly stringent (currently requiring a human safety driver) but has a less elaborate permanent framework.
Conceptual positioning of reviewed states according to relative regulatory elaboration and stringency
  • Global Convergence: While the U.S. remains decentralized and industry-led, the rest of the world is converging toward "type-approval" models. These involve scenario-based testing, ODD-specific (Operational Design Domain) certification, and mandatory "safety cases."

Key Results

The study highlights the stark differences in how states handle evidence and oversight.

  • California (Tier 3): Operates the most complex program, differentiating permits by operational phase (testing vs. deployment) and vehicle type. It involves two agencies: the DMV for technology and the CPUC for passenger service.
  • Texas (Tier 3): Recently moved to a permit-based model (effective late 2025) in response to increased deployment activity, though its requirements remain less prescriptive than California’s.
  • North Carolina (Tier 1): Currently maintains a "light touch" approach, which minimizes administrative burden but offers limited proactive oversight as driverless operations expand.

The authors provide a detailed breakdown of these requirements in their analytical framework:

Questions Used to Review State AV Certification Practices

Limitations and Future Work

The authors acknowledge that their typology is a snapshot of a rapidly evolving field; state legislation can change overnight. Furthermore, while international models offer "instructive context," the U.S. federal-state division of authority makes direct transferability difficult.

Future Directions: * Workforce Expertise: States moving toward Tier 2 or 3 will need significant new expertise in ADS safety concepts, cybersecurity, and simulation-based scenario analysis. * Infrastructure Integration: Future research should look at how state governance must evolve to include "machine-readable" infrastructure, such as digital representations of work zones and signal timing. * Mixed-Fleet Management: As AVs operate alongside human drivers, states will need to update driver education and incident response protocols for "mixed-fleet" realities.

Summary updated

Abstract

Automated vehicle (AV) deployment in the United States is expanding without a unified federal certification framework, leading to substantial variation in how states authorize and oversee AV testing and deployment. This paper examines how U.S. states regulate Society of Automotive Engineers (SAE) Level 4-5 automated driving systems (ADS), identifies common certification approaches, compares them with international practices, and evaluates the implications for North Carolina's future regulatory strategy. Publicly available state statutes, administrative regulations, agency guidance, and federal policy documents were reviewed. Ten states, selected to capture the range of regulatory approaches, were analyzed using a structured framework covering operational scope, governance and oversight, federal interaction, and implementation outcomes. Certification practices in the European Union, China, Japan, and Australia were reviewed for comparative context. Three state certification models were identified, reflecting increasing regulatory involvement: baseline legal compliance, administrative self-certification with ongoing oversight, and permit-based authorization requiring formal agency review before deployment. North Carolina currently aligns with the baseline compliance model. International frameworks are converging toward scenario-based testing, operational-design-domain-specific approval, safety-case documentation, cybersecurity assurance, and lifecycle oversight, although direct transferability to the U.S. is limited. The paper develops a three-model typology for comparing state AV certification approaches and integrates state practices with international frameworks, providing a consistent basis for evaluating differences in regulatory involvement, governance complexity, and institutional responsibilities across jurisdictions.
BibTeX
@misc{ahsan2026national,
  author = {Sk Nahia Ahsan and Md Abdullah Al Hasan and Danjue Chen and George List},
  title = {{National and State-Level Approaches to Automated Vehicle Certification: A Comparative Framework for State Operational Governance}},
  year = {2026},
  eprint = {2611.08681},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.08681},
}
Figure