FAAAA preemption was the wall every freight broker stood behind. The Supreme Court took it down on May 14, 2026 in Montgomery v. Caribe Transport II, LLC.
The reporting covered the ruling. It didn't cover what brokers are supposed to do about it. Removing the federal preemption shield exposed the duty — it didn't define what reasonable carrier vetting looks like. That gap is where most brokerages sit exposed right now: not because they do nothing, but because what they do lives in dispatchers' heads, not in a form that survives scrutiny.
This is the operational definition. What gets checked, when, by whom, and what the paper trail needs to look like.
The three layers of defensible vetting
Documentation is not a single event at onboarding. The ordinary-care standard implies a process — one that runs before a carrier is approved, before each load is tendered, and on a recurring basis between bookings. Three distinct layers.
Layer 1: Carrier onboarding
Before a carrier moves freight, the record should capture:
Authority and registration
- Active operating authority — active status at time of approval, not just existence
- USDOT number verified against the FMCSA carrier database
- Common or contract authority confirmed for the freight type
Safety record
- SMS BASIC percentiles across all seven categories at time of approval
- Any Unsatisfactory or Conditional safety rating — and if approved despite one, the written rationale
- CSA intervention history (warning letters, cooperative safety plans, notices of violation)
- Crash and inspection history — not just the score, the pattern
Insurance
- Certificate of insurance verified directly with the issuing agency — not a copy forwarded by the carrier
- Cargo coverage confirmed for the commodity type and load value
- Policy expiration logged with a re-verification trigger date
Identity
- Physical address confirmed — not a mail drop or virtual office
- Ownership structure — particularly relevant for carriers recently transferred, rebranded, or operating under related authorities
The output of Layer 1 is a carrier file: what was checked, when, what the values were, who approved it. Not a checkbox.
Layer 2: Load-level documentation
Carrier approval does not mean carrier safety is static. Between an onboarding file and a load two years later, insurance can lapse, authority can be revoked, percentiles can shift materially.
Before tendering each load, the record should capture:
- Authority status re-verified at time of booking — not "we approved them last year"
- Insurance active and covering this load's commodity and value — verified, not assumed
- No open out-of-service order since the last booking
- If the carrier has a Conditional rating at tender: the written rationale for proceeding
This does not require rebuilding the carrier file every time. It requires a documented re-check that timestamps who checked what and when.
The pattern that fails in litigation: a carrier's authority was revoked three months before the incident load, but the record shows only the original approval. The inference is negligence. The counter-argument — "we check before every load" — is not usable without the documentation to prove it.
Layer 3: Periodic re-screening
FMCSA updates SMS BASIC percentiles monthly. Insurance policies lapse, sometimes without notice. Authority gets revoked on carriers still sitting in routing guides. A carrier that passes onboarding in January can be materially different by October.
A defensible cadence:
- Monthly: flag any active carrier whose percentiles have crossed a threshold since the last booking
- Quarterly: full re-verification for carriers used heavily in the prior quarter — authority, insurance, rating
- Triggered: any carrier involved in a claim, out-of-service order, or adverse review — full re-screen before next use
The re-screening does not need to be manual for every carrier. It needs to be systematic and logged. A process that runs and leaves no record is a process that, legally, did not run.
What "documented" actually means
Documentation is not a database field that says "approved." It is a trail that answers four questions a plaintiff attorney will ask:
- What information did you have about this carrier when you selected them?
- When did you last verify it before this specific load?
- Who made the decision, and on what basis?
- What would have caused you not to use this carrier?
If your system cannot answer all four — for any load, on any date — the standard is not met. The practical minimum: a carrier record with timestamped entries, a per-load verification log, and a written SOP that describes the process and confirms the team was trained on it.
The override problem
The highest-risk failure is not the absence of a process — it is undocumented overrides of the process.
Dispatcher needed a truck. The preferred carrier wasn't available. They found one with a Conditional rating and moved the load. No record of why.
In litigation, that load looks identical to a reckless booking. The documented override — carrier unavailable, alternate selected, Conditional rating noted, shipper approval obtained, no insurance gap — is the difference between a defensible exception and an indefensible pattern. Override documentation is not about tracking mistakes. It is about distinguishing operational decisions from negligence.
What reasonable looks like
The standard is ordinary care — not perfection. Brokers are not liable for every carrier crash. They are exposed when their selection process falls below what a reasonable broker would do, and they cannot demonstrate otherwise.
The practical threshold: a reasonable broker checks authority, insurance, and FMCSA safety data before each load and keeps records that show it. Brokers who cannot demonstrate this — not because they failed to do it, but because nothing was logged — face the same inference as brokers who did nothing. Documentation is not evidence that you did your job. It is your job.
If you have no documentation today
The process does not need to be built from scratch. Most brokerages already have carrier approval workflows in their TMS, dispatcher habits, and informal standards. The gap is usually not practice — it is the paper trail.
Starting points:
- Export your carrier list. Flag the last booking date for each. Any carrier used in the last 90 days needs a current file.
- Pull insurance certificates. Verify your top carriers by volume against their certificates today. Note lapses and upcoming expirations.
- Write down what your dispatchers actually do. The informal process, converted to a written SOP, beats no SOP.
- Create a load log. Even a spreadsheet: carrier, authority verified, insurance verified, date, dispatcher. Build the habit before building the system.
The window is this quarter. Cases paused on preemption grounds are reactivating. Underwriters are tightening standards on contingent and broker liability coverage — and documentation of a vetting process is increasingly what separates a clean renewal from a repriced one. The documentation does not take a year to build. The discipline to maintain it is the harder part, and that requires a process, not a good intention.