Beyond the TSA Checkpoint: Why Mobile Driver's License Rollout Is a State Identity Infrastructure Decision
Back to Signal
State & LocalCybersecurityGovtechGovernmentModernization

Beyond the TSA Checkpoint: Why Mobile Driver's License Rollout Is a State Identity Infrastructure Decision

November 19, 2026Jess Loban

ISO 18013-5 and the mDL Technical Baseline

The governing international standard is ISO/IEC 18013-5:2021, published in September 2021, which TSA describes in its 2023 proposed rule as critical to interoperability, security and privacy for federal acceptance. The parts of the standard that matter for this decision fall into two layers.

How a verifier reads the credential. TSA's proposed rule describes two retrieval modes:

  • Device retrieval (offline): the reader retrieves data directly from the holder's phone. NIST's June 2026 overview of credential presentment explains that the connection is started by NFC or a QR code, and the data then moves over Bluetooth Low Energy, NFC or Wi-Fi Aware. The QR code is how the session begins, not a separate mode.
  • Server retrieval (online): the reader retrieves the data from the state driver's licensing agency. Commenters raised concerns that this mode lets mDL use be tracked, and TSA concluded that server retrieval was not appropriate for federal acceptance at that time.

What the holder shares. NIST notes that in an age check the wallet can answer with only the minimal attribute required, such as a yes or no, rather than the full record. Whether a given relying party can use that capability depends on which data elements the state put in the credential and which requests the state's governance permits.

A third piece now matters for state agencies. ISO/IEC 18013-7, which NIST describes as a newer specification building on 18013-5, adds presentation of an mDL over the internet, using OpenID for Verifiable Presentations or the W3C Digital Credentials API. TSA's 2024 final rule called this an upcoming specification for "unattended" verification and stated that unattended verification is not currently a REAL ID use case. In other words, the federal baseline states are building to is in-person only.

The TSA Optimization Problem

TSA now accepts digital IDs at more than 250 airports, presented through Apple Wallet, Google Wallet, Samsung Wallet or a state app, and it lists 22 states plus Puerto Rico as participating. Travelers must still carry a physical ID. Under the October 2024 final rule, effective November 25, 2024, federal agencies may accept a state's mDL only if TSA has granted that state a temporary waiver.

The waiver criteria are the closest thing to a published specification for a "TSA-compatible" mDL. Among them, a state must:

  • Be in full compliance with REAL ID and issue mDLs based on a REAL ID-compliant physical card
  • Issue mDLs interoperable with ISO/IEC 18013-5:2021 and use the AAMVA mDL data element set, including the DHS_compliance field agencies check
  • Include the standard's mandatory data elements, among them name, birth date, issue and expiry dates, issuing authority, document number and portrait, plus resident address, sex, portrait capture date and signature
  • Meet detailed security requirements for its certificate authority and issuance systems, including certificate revocation

On the verifier side, 6 CFR 37.8 requires an accepting federal agency to use an mDL reader to retrieve and validate the data as ISO/IEC 18013-5 requires. The final rule states that mDLs must work in an offline environment.

That is a coherent design for an attended checkpoint run by an agency that is not the issuer. It is not a design for state agencies. Nothing in the waiver criteria requires a state to offer a verification service its own agencies can call, to support remote presentation, or to define which agencies may request which data and under what authority. A state that builds exactly to the waiver produces a credential that works at airports. It does not automatically produce a credential that works for state agency identity assertions.

What State Agencies Actually Need

State programs verify identity at specific, high-stakes points, and the requirements differ from a checkpoint in ways that matter.

Benefits enrollment: Medicaid, SNAP, unemployment insurance and child care subsidy programs verify identity to establish eligibility and prevent fraudulent enrollment. An agency relying on an mDL for that purpose needs to show that the credential belongs to the applicant, that it was valid when presented, and that the decision can be defended in an administrative hearing or audit.

Professional licensing: occupational licensing boards for nurses, contractors, teachers and engineers verify identity at initial licensure and renewal, and may need to link the license record to a persistent identity across renewals and name changes.

Hearings and redeterminations: agencies that conduct hearings, appeals and benefit redeterminations need reliable identity verification for participants who are often not physically present. In-person device retrieval does not cover that case. Remote presentation of the kind 18013-7 defines would, but only if the state's wallet, the agency's systems and the governance framework support it.

The common thread is current status. Device retrieval lets a verifier confirm that the data was signed by the issuing state and has not been altered. Whether it also reflects a DMV record change made after the credential was provisioned depends on choices the state makes about credential validity periods, refresh and status checking. A benefits agency should know the answer before it accepts mDL assertions, and in a DMV-only program it is often nobody's job to publish that answer to other agencies.

Who Is at the Table

When an mDL program is chartered as a DMV project, the design table tends to include DMV technology and policy staff, wallet and app vendors, and the state IT governance office to varying degrees. TSA's requirements are present through the waiver criteria. The agencies that will eventually want to rely on the credential, such as health and human services, labor and licensing boards, may not be. The state CISO may not be either, even though revocation, key management and the risk of creating a single point of failure for state identity are security questions.

The result is a credential designed around the use cases represented in that room. This is not a failure of DMV program management. The DMV is solving the problem it was handed: get the credential accepted at airports and by in-person verifiers such as retailers checking age, which drives adoption and justifies the investment. The gap is that no one is tasked with the broader-purpose credential alongside the TSA deployment, so the TSA deployment becomes the architecture by default.

The Technical Decisions That Determine Usefulness

Four decisions determine whether an mDL ends up narrow or broad.

Verification and status service: does the state offer other agencies, under a governance agreement, a way to confirm the current status of a credential? Without one, an agency can validate what is presented but cannot see later changes. Server retrieval is one way to provide this, but it carries the tracking concerns TSA cited, so the design should limit what the issuer learns about each use.

Data elements and request rules: a waiver-compliant credential already carries a document number and address alongside name, birth date and portrait, which covers basic record matching. The harder question is governance: which agencies may request which elements, and whether any program needs something the credential does not hold. Adding elements later generally means reprovisioning credentials.

Identity binding for remote use: portrait comparison works at an in-person checkpoint. Remote enrollment needs a different approach, such as liveness detection, comparison against existing state records, or identity proofing aligned with NIST SP 800-63-4, the digital identity guidelines NIST finalized in July 2025.

Legal authority and audit: who may request credential attributes, under what authority, and with what audit trail? A state that launches without a cross-agency data use framework may find that agency counsel will not approve reliance on DMV-issued data outside the DMV context. That legal work can take longer than the technical work.

A Credible Alternative

A state that decides to build mDL explicitly as identity infrastructure, rather than as a DMV modernization project, takes a different approach from the start.

  1. Widen the charter. Convene the state CIO, state CISO, health and human services and Medicaid leadership, the labor department CIO, and counsel for each major program expected to accept mDL assertions. Define those use cases before selecting technical options.
  2. Plan the online capability now. The TSA-facing deployment can use device retrieval as the waiver requires. The architecture should still include the status service and the remote presentation path agencies will need, even if few agencies use them at first. Adding them after launch means reworking issuance and certificate infrastructure around a credential already in residents' phones.
  3. Write the acceptance framework before launch. Define each agency's legal authority to accept mDL assertions, which attributes each program may request, and what audit and retention rules apply. Without it, the credential exists but agencies may not be able to use it.

The case against this approach is budget and timeline: a broader design process takes longer and costs more than a TSA-only launch. That alternative makes sense when the immediate risk is losing federal acceptance or when no agency has a near-term use case. Its weakness is that a deferred second phase competes with every other priority once the initial deployment is working and nothing has visibly failed, so the state should decide explicitly whether it is accepting that risk.

What Changes the Analysis

If a state has already deployed a TSA-focused mDL, the question is whether the architecture can be extended. Where the backend can already support status checks or remote presentation but was never opened to other agencies, the remaining work is mostly governance and integration. Where the credential's data elements, wallet support or key management would have to change, extension may require reprovisioning.

The federal baseline can also move. TSA's waiver rule is explicitly a bridge to a future "Phase 2" rulemaking with comprehensive mDL requirements; a search of the Federal Register on November 19, 2026, found no Phase 2 proposal yet. Waivers are not permanent either: TSA's participating-states page says it will stop accepting Utah's mDL on January 1, 2027, and West Virginia's on May 4, 2027. A state designing beyond the waiver should track both.

A state evaluating its current mDL posture should answer four questions:

  1. Can agencies other than the DMV confirm a credential's current status, and through what service?
  2. Which data elements does the credential contain, and which agencies are authorized to request each one?
  3. What is the actual delay between a DMV record change and a verifier seeing it, under both device retrieval and any online path?
  4. Has any agency outside the DMV piloted mDL-based identity verification, in person or remotely, and if not, is the obstacle technical or legal?

If the answers reveal a TSA-only design, the near-term decision is whether to extend the current system or to use a planned reprovisioning cycle to issue a broader-purpose credential. Physical card renewal is not a fast lever: 6 CFR 37.5 allows REAL ID cards to remain valid for up to eight years, so the mDL's own provisioning and refresh cycle is the more practical opportunity.

That window is not unlimited. States that begin the broader design process now, before the first deployment hardens into legacy infrastructure, have more options than states that wait for a benefits program to discover that the credential the DMV built cannot serve its needs.

Sources and Further Reading

Spartan X's published state work spans AI strategy, cybersecurity architecture advisory and systems integration across multiple agencies in Colorado and Florida. That cross-agency vantage point is what an mDL program needs at the design stage: someone who can translate a benefits agency's eligibility and audit needs, a CISO's revocation and key-management concerns, and counsel's data-use questions into architecture decisions before the DMV's launch locks them in.

Share this article
LinkedIn

BUILD WITH US

Ready to Solve Hard Problems?

Spartan X builds AI systems, autonomous platforms, and cybersecurity solutions for defense and national security.