What actually happened
On 14 May 2026 the Supreme Court decided Montgomery v. Caribe Transport II, holding that state-law negligent-selection claims against freight brokers are not preempted by the Federal Aviation Administration Authorization Act. Brokers can now face those claims in state court in circumstances where a preemption defence previously ended the matter early.
We are an operations company, not a law firm, and nothing here is legal advice. Your counsel should tell you what the ruling means for your risk. What we can talk about is the operational consequence, because that is where the work lands.
The operational consequence
If a carrier selection decision can be examined years later, the question stops being "did you have a policy?" and becomes "can you show what you did on this particular carrier, on this particular day?" Those are very different requirements. The first needs a document. The second needs a discipline.
A vetting policy nobody can evidence is, in practice, a vetting policy nobody followed.
Most brokers we speak to already have a sound checklist. What they do not always have is proof that the checklist ran identically on the Tuesday afternoon when dispatch was three loads behind and one person was covering two desks.
What a defensible file contains
- The check performed, named the same way every time
- The source consulted, with a screenshot or export attached
- The timestamp, and who performed the check
- The result, including the ones that passed unremarkably
- The exception path: what failed, who was told, what they decided
That last line matters more than people expect. A file showing that a problem was found, escalated and knowingly accepted by an authorised person tells a much better story than a file with no record of the problem at all.
Why this work suits a supervised desk
Vetting is repetitive, rule-bound and exacting, and it has to happen at the least convenient moments. Those are exactly the conditions under which an in-house team, pressed for time, starts taking shortcuts that nobody intends and nobody notices.
A dedicated desk removes the time pressure from the checklist. The people doing it have no competing dispatch duty, they run the same steps in the same order every time, and a QA reviewer samples their files daily. The broker keeps every approval decision; the desk keeps the record.
Where to start
Write the checklist down as a numbered procedure, including the escalation path. Decide what evidence each step needs. Then run it on fifty carriers and see whether the files look the same. If they do not, the problem is capacity, and capacity is a solvable problem.