I've sat in on enough deposition prep sessions to know what breaks a defense. It's almost never the absence of documents. Brokers have documents. They've got a signed carrier packet, a COI on file, a SAFER printout from the day they first approved the carrier. They walk into the room with a folder that looks thorough. And then the plaintiff's attorney picks up the SAFER printout and asks one question: "Can you tell me the date this was run?"
That's when the room gets quiet.
The date on the printout is fourteen months before the crash. The crash happened on load 62 with that carrier. Between the approval and the accident, nobody checked again. And now you're being asked to explain, under oath, why a printout from over a year ago demonstrates you met your duty of care on last March's load.
After Montgomery v. Caribe Transport II — the Supreme Court's May 2026 ruling that the FAAAA does not preempt state-law negligent-selection claims against freight brokers — that question isn't hypothetical anymore. It's the question in every case where the plaintiff can show you kept using a carrier without ever verifying that the carrier was still safe to use. The duty to vet isn't a one-time checkbox. It's a recurring obligation. And the way you prove you met it isn't by having a file. It's by having a file where every entry is tied to a date.
What "due diligence" actually means in discovery
Most brokers interpret due diligence as a carrier approval process. You check the MC, verify the authority, pull the SAFER snapshot, collect insurance, get the carrier packet signed, and approve them. That's the front door. That's not what gets tested when there's a $2.3M wrongful-death claim on the table.
What gets tested is the transaction. Not the approval — the specific load that's at the center of the lawsuit.
49 CFR § 371.3 says brokers must maintain records "of each shipment," including the identity of the carrier used. It's a recordkeeping requirement with a three-year retention window. What § 371.3 doesn't say — and what a lot of brokers incorrectly infer — is that a single carrier approval record satisfies the per-transaction documentation obligation. The reg tells you what to keep. It says nothing about how current your safety verification needs to be for any given load.
That's where Montgomery changes the analysis. State tort law does. And plaintiffs' attorneys arguing negligent selection are going to argue that approving a carrier once and then tendering loads indefinitely, without ever rechecking their safety status, doesn't satisfy a broker's duty of care. You can argue about what the standard should be. But if the carrier's Unsafe Driving BASIC crossed the intervention threshold six months before the crash and you have no record showing you checked their SAFER in the weeks before that load, you're going to have a bad day explaining why that's irrelevant.
The scenario that's easier to ignore than it should be
Take MC-1249876, DOT-3571034. A four-truck dry-van operation, approved by a broker in early 2025. At approval: clean record, authority almost two years old, no crashes in the SAFER snapshot, OOS rate under 5%, Unsafe Driving BASIC at 42. Good carrier. Approved, filed, added to the load board.
By September 2025, the Unsafe Driving BASIC had climbed to 77. Above the 65-point intervention threshold. The OOS rate was at 18.3% on vehicle inspections. Two of the SAFER violations were brake-related. None of that showed up in the broker's file because nobody had looked since the original approval.
Load 52 went out in October. Load 53 in November. Load 62 in January, with a flatbed hauling $180K of fabricated steel. On I-90, the carrier's truck rear-ended a passenger vehicle stopped for construction traffic. The driver had a brake malfunction. The passenger car's driver was critically injured. The subsequent litigation named the broker as a defendant for negligent carrier selection.
The broker's defense: "We approved this carrier. We have the documentation." The plaintiff's rebuttal: "Your documentation is from fourteen months ago. In the fourteen months since you approved this carrier, their Unsafe Driving BASIC crossed the intervention threshold, their OOS rate tripled, and you have no record showing you verified any of that before Load 62."
There's no good answer to that rebuttal. Not because the broker was reckless — they approved a carrier that looked clean. But because their file told the story of the approval, not the story of the ongoing relationship.
The timestamp problem is structural
Here's why this happens: carrier approval is a workflow. Load tendering is a different workflow. Most brokers have built processes around carrier approval, not around per-load safety verification. And that made sense before Montgomery changed the litigation landscape.
The problem is that SAFER data isn't static. BASIC percentiles update monthly. Insurance status changes. OOS rates shift with each new inspection cycle. A carrier that was solid at approval might be showing warning signs six months later. Continuous monitoring exists precisely because a snapshot from fourteen months ago tells you nothing about the carrier's current safety posture.
But monitoring without documentation solves half the problem. If you're using a monitoring service that alerts you when a BASIC crosses a threshold, that's valuable. If those alerts aren't tied to load records — if you can't show which loads moved after an alert fired or prove you reviewed the carrier's status before tendering — you've got monitoring without a defense.
The timestamp is the defense. The document is just a prop. A printout in a folder proves you can operate a printer. A timestamped record tied to a load proves you verified a specific carrier's safety status before you tendered that specific load.
What the file should actually look like
If a deposition required you to demonstrate your diligence on any specific load, you should be able to pull the carrier's record for that load and show:
What you verified, when you verified it. The safety check needs to be contemporaneous with the load, not with the approval. "We pulled their SAFER before the load tendered" is a statement you need a record to back up. That means the check has to generate something with a timestamp, and that record has to be retrievable.
What the safety data showed at that moment. This matters because if the carrier had a BASIC above the intervention threshold and you tendered anyway, you need to show that you saw the data and made an informed decision — not that you didn't see it because you never checked. Awareness of a risk that you evaluated and accepted is defensible. Ignorance of a risk that would have shown up if you'd looked is not.
A consistent record across loads. One timestamped check before one load isn't a process. A consistent record of per-load verification with the same carrier demonstrates a practice. Defense attorneys argue practice; plaintiff's attorneys attack exceptions. Don't give them exceptions.
The carrier's FMCSA authority status. 49 CFR § 387.303 sets the minimum financial responsibility requirements — $750K for general freight under 10,001 lbs, $1M BIPD for many hazmat movements. Verifying authority active and insurance at or above these thresholds should be part of the per-load record, not just the approval record, because authority can be revoked and insurance can lapse without any warning to the brokers who last checked months ago.
How I document this
When a load goes on the board with a carrier I've used before, the carrier's safety data gets pulled at the time of booking, not pulled from the last approval record. That means:
- Current FMCSA authority status (active, not revoked or suspended)
- Current insurance on file with FMCSA, verified in the L&I system, not assumed from last month's COI
- SAFER BASIC percentiles current as of the most recent SMS update
- OOS rate (vehicles and drivers) compared to the national average
- Any crash incidents recorded since the last review
All of that gets attached to the shipment record with a timestamp. Not a separate "carrier file" that lives in a shared drive somewhere — attached to the specific load record so the connection between the verification and the transaction is explicit.
DOTScreener generates a Carrier Selection Record tied to each screen. When a carrier's status changes — BASIC crosses a threshold, insurance status changes, authority gets flagged — the monitoring layer catches it and the record reflects what was known at the time of each load. The re-filed record is the per-transaction documentation. That's what you want to be able to produce in discovery: not a folder of general approvals, but a trail of specific checks, each tied to the load it covered, each timestamped.
The standard isn't perfection. Nobody expects a broker to predict a brake failure. What the standard requires, post-Montgomery, is that you exercised reasonable care at the time of each tender — that you looked at what was available to look at, documented what you found, and made a defensible decision. A static carrier file doesn't demonstrate that. A per-load record with timestamps does.
Your best defense in a negligent-selection case isn't a good lawyer. It's a file that makes the plaintiff's attorney's job difficult. Build that file before you need it.
How I document this
Carrier name, MC/DOT, load number, date of screen (timestamped). FMCSA authority status at time of screen. Insurance verification — carrier, policy number, coverage limits, verified against L&I records, not just a COI on file. BASIC percentiles pulled at the time of booking (not from last month's summary email). OOS rates, driver and vehicle. Any SAFER crash entries added since the last review. Decision: approved or flagged, and if flagged, the reason documented. If approved despite a flag (above-threshold BASIC with no other red flags), the rationale recorded.
If the carrier's record changes between booking and pickup, I want to know. That's what continuous monitoring is for. And if I get an alert, that alert and my response to it go in the load record too.
Fourteen months between checks isn't diligence. It's a folder.
— Mason Lavallet
Founder, DOTScreener.com
Automate your carrier vetting
DOTScreener runs every check in this article automatically — live FMCSA data, documented decisions, tamper-evident audit trail.
Go deeper
Related Articles
The Carrier SAFER Can't Find
FMCSA moved carrier registration to its Motus system in May 2026 and started issuing randomized 8+ digit MC numbers that legacy lookup tools—including SAFER's MC search—can't resolve. Here's what that means for your vetting process and your paper trail.
Broker GuidesFMCSA's Intervention Threshold Isn't a Clearance. It's a Priority Queue.
A carrier can sit above FMCSA's intervention threshold for months before anyone investigates them. That doesn't make them safe to book — it means you're the one doing the monitoring until FMCSA gets there.
Broker GuidesThe Insurance Lapse Your SAFER Pull Won't Show You
A one-time SAFER pull is a photograph, not a film — and plaintiffs' attorneys know how to find the frames you missed. Here's what insurance lapses look like in discovery and why a single check-at-booking isn't enough anymore.