The Supreme Court's Montgomery ruling removed the preemption shield from negligent-selection claims and then, deliberately, stopped. The Court did not define what reasonable care in carrier selection looks like. That definition will now get built the expensive way — case by case, deposition by deposition, out of the records of actual brokerages.
Which makes this the practical question of 2026 for every broker desk: when a plaintiff's lawyer reconstructs the day you selected a carrier, what will the record show?
Think like the discovery request
A negligent-selection case runs on documents, and the requests are predictable. Your written carrier-selection policy, if one exists. The onboarding file for the carrier involved. What safety data you pulled and when. Who approved the carrier and against what threshold. Whether your stated process was actually followed on this load — or waived because it was 4:50 on a Friday.
Every gap between what your policy says and what your records show becomes the plaintiff's theme. Worse is having no policy at all, because then the standard of care gets defined entirely by the other side's expert.
The record that holds up
It has four properties, none of them exotic.
Defined. A written procedure: which checks, which data sources, which thresholds fail a carrier. A defined bar you consistently apply beats a high bar you sometimes apply, because consistency is what "procedure" means to a jury.
Grounded in available data. FMCSA authority and history, inspection and out-of-service performance, crash exposure, insurance in force, identity signals. After Montgomery, "the data was public and you didn't look" is the sentence to fear. Kavanaugh's concurrence pointed the same direction from the defense side: reasonable procedures, reputable carriers.
Dated and per-carrier. Not "we generally check safety scores" — a record that this carrier was checked on this date and showed these values. Vetting that isn't written down will be treated, in a courtroom, as vetting that didn't happen.
Alive. A carrier vetted in March and never looked at again is a March fact. Insurance cancels mid-term; safety scores move; authorities change hands. Continuous monitoring — with alerts logged — shows care that continued past onboarding, which is where the next wave of claims (negligent retention) is already pointing.
The honest cost argument
Ninety percent of this is data lookup and record-keeping — exactly what should be automated. When Sentinel vets a carrier in Dollar Traq, every check runs against our own FMCSA warehouse and lands in the carrier's file with its source and timestamp; monitoring keeps the file current after onboarding. The by-product is the thing Montgomery just made valuable: a defensible, dated, per-carrier evidence trail that exists because the work happened, not because someone remembered to document it.
Brokers spent years arguing they shouldn't be judged on carrier selection. That argument is over. The next one — we selected carefully, and here's the proof — is winnable. But only on paper you built before the crash.
General information, not legal advice — build your retention and policy details with transportation counsel.