You Screened the Carrier. The Screen Came Back Yellow. Now What?

Six months before the crash, a colleague of mine ran a screen on the carrier. The screen came back with flags: HOS Compliance BASIC at the 74th percentile, a Level 2 OOS vehicle violation in the prior 12 months, nine months of authority. Not a FAIL. Not a clean green. A WARN. He noted the result in an email to himself, talked to his ops lead about it for about four minutes, and approved the carrier anyway. No written rationale. No documented mitigation. No escalation chain. He used them once, went fine, added them to the approved list, and moved on.

Eight months after that first load, that same carrier ran a red light in rural Missouri. A driver in a pickup truck died. Plaintiff's counsel subpoenaed his carrier file the following week.

The lawyer didn't need to prove my colleague acted in bad faith. All they needed was the WARN result with no documentation behind it. They had a broker who knew something was flagged, had no written reason for proceeding, and — according to the file — made the decision in silence.

That's where most brokers are. Not on the clean approvals. Not on the hard declines. On the yellow in the middle.

The WARN case is the hardest one to defend

When your screen comes back clean, the documentation is easy. You ran the check, everything cleared, you approved, you moved on. When it comes back as a hard FAIL — the carrier's OOS, authority revoked, insurance lapsed — you don't use them, and there's nothing to defend.

The WARN case is different. It says: here is a carrier with issues, and you chose to use them anyway. The question the plaintiff's lawyer will ask is: what was your process for making that choice?

If your answer is "I thought about it," that's not a defense. It's an invitation.

Most brokerage operations I know don't have a written policy for the WARN. They have a general vetting standard — check authority, check insurance, check SAFER, maybe run a BASIC screen — and they have a rough sense that carriers with bad scores or active OOS don't get approved. But the in-between? That lives in people's heads. And heads are not admissible in discovery.

What a WARN actually means operationally

A WARN doesn't mean the carrier is unsafe. It means your screening criteria flagged something that requires a decision, not just a process step. The decision might be: we proceed with this carrier, but we're taking the following additional steps. Or: we decline. Or: we escalate to a compliance lead before approving.

All three of those are defensible. None of them are defensible without a written record.

Here's a scenario. You're booking a dry van load of industrial parts, $94,000 declared, shipper out of Chattanooga, delivery in Columbus. You run a screen on the carrier: MC-1847293, DOT-4192087. Eight months of authority. Clean insurance. No crashes in SAFER. Fourteen inspections, OOS rate at 7.1%. BASIC scores mostly clean, but HOS Compliance at the 71st percentile — above FMCSA's 65% intervention threshold. Screen result: WARN.

The HOS percentile is the flag. The carrier's drivers have a pattern of hours-of-service violations that puts them in the top third for their peer group. That's not a disqualifier per se, but it's a signal. The question is: what do you do with it?

Option A: You approve the carrier with no notes and tender the load. That's what most people do. It's the option that ends badly in a deposition.

Option B: You have a written policy that says what a WARN requires. You follow it. You document it.

Option B is what you need.

What the regulation says — and doesn't say

49 CFR Part 385 defines the FMCSA's safety fitness determination framework. The agency uses BASIC percentiles to identify carriers for intervention, with 65% (or 80% for the Crash Indicator BASIC) as the threshold above which FMCSA starts taking notice. But Part 385 doesn't tell brokers what to do with a carrier above that line. The federal standard for broker conduct is still the emerging "reasonable care" standard that Montgomery v. Caribe Transport II, LLC brought into focus.

What Justice Barrett wrote in the unanimous opinion — decided May 14, 2026 — is that the FAAAA does not preempt state-law negligent-selection claims. The gate is open for plaintiffs to argue that a broker failed to take reasonable care in selecting a carrier. What "reasonable care" requires is now a question for juries in state court.

That means your WARN policy IS your standard of care. If you have a written policy that says "any WARN result requires documented escalation and approval with a written rationale before the carrier can be used," and you follow it, you can explain your decision-making. If you have no policy, your standard of care is whatever you happen to do — and that is not a position you want to defend.

The FMCSA has told you, through the BASIC system, that a carrier above the intervention threshold has been flagged for attention. When you knowingly use a carrier above that line and can't explain why — in writing, with documentation — you've handed the plaintiff's lawyer a narrative: "The company that makes the safety rules flagged your carrier, and the broker knew it and did nothing."

What a written WARN policy looks like

It doesn't have to be a twelve-page compliance document. It should answer four questions clearly:

Who decides? A WARN shouldn't be a solo call by the booking rep. It should require sign-off from a compliance lead or a supervisor. The decision needs an owner who isn't just the person making the commission.

What did you see? The policy exception record needs to document the specific screen result — which BASIC was flagged, at what percentile, what other factors were visible at the time. That snapshot matters because it establishes what you knew.

Why did you proceed? One paragraph is enough, but it has to exist. "Carrier has nine months of authority and the HOS percentile is elevated, but inspection frequency is high (fourteen inspections in 12 months), OOS rate is within acceptable range at 7.1%, and the shipper's lane and load type don't present elevated fatigue-risk factors." That's a defensible rationale. "Seemed fine" is not.

What additional steps did you take? This is where you list the mitigations. Did you add a T-call requirement before pickup? Did you require a certificate of insurance current within 30 days rather than accepting the file copy? Did you add a contract term requiring the carrier to confirm driver hours availability before acceptance? Document what you did differently because of the flag.

None of this takes more than ten minutes per exception. The value isn't in the time — it's in the paper.

The quiet problem: carriers stay on approved lists after a WARN

Here's what makes this worse. Most brokers don't re-vet carriers at tender. They vet at onboarding. Once a carrier is on the approved list, they get used again and again without the BASIC scores being re-run, without the SAFER snapshot being refreshed, without anyone asking whether the WARN from the original onboarding ever got resolved.

A carrier can be onboarded with a WARN, approved via a policy exception, and then — as their HOS Compliance BASIC climbs to the 84th percentile over the following year — never flagged again until a load goes wrong.

Continuous monitoring is supposed to catch this. But monitoring alerts on specific changes — an insurance lapse, an OOS event, a BASIC moving into alert territory — not on the accumulation of risk that happens gradually. That's where re-filed Carrier Selection Records come in. If you're re-running a screen at each tender, you catch the carrier whose WARN from onboarding has quietly become a different risk profile. That's not onboarding vetting. That's per-load vetting. And those two things are different in front of a jury.

When you run a screen in DOTScreener and the result comes back WARN, the platform captures the decision workflow: who approved the exception, what they documented, what the screen showed at that moment. That record re-files with the load. If a plaintiff's attorney subpoenas your carrier file eighteen months later, you have a timestamped policy exception note attached to the screen result, not a memory.

That's not a product pitch. That's the evidentiary standard that a post-Montgomery world requires.

The carriers where a WARN is almost always worth a closer look

Not every WARN is equal. A few categories where the flag should get more weight:

HOS Compliance above 75th percentile on a long-haul load. The BASIC is already elevated — adding miles to the haul increases the fatigue-risk profile of what you're already seeing in the data.

New authority (under 12 months) with any BASIC above 65%. Short authority with elevated scores means you're extrapolating from a small inspection sample. The score could improve — or the sample could be masking a worse picture.

Vehicle Maintenance above 70th percentile on a heavy or oversized load. Equipment failure risk on a flatbed pulling steel coil is a different problem than a dry van pulling apparel. The BASIC has more consequence when the load type amplifies it.

Any BASIC above the intervention threshold combined with a Conditional or "Not Rated" safety fitness designation. A "Not Rated" carrier with elevated BASIC scores hasn't had a formal compliance review. You're the first line of scrutiny.

In those situations, the WARN policy exception gets higher scrutiny before the exception is approved — or the load gets declined. Either answer is defensible. What isn't defensible is approving without a record.

How I document this

When a screen returns WARN, I capture the following before any approval decision is made:

A screenshot of the screen result with the timestamp visible — which BASICs flagged, at what percentile, and the date the screen ran. I keep this in the carrier's file, not in a separate spreadsheet.

A written note from the approving person (not the booking rep) that answers the four questions above: who decided, what they saw, why they proceeded, what extra steps were taken. One paragraph is enough. It has to be in the file dated before the load tender.

A notation on the rate confirmation that the load is subject to additional verification steps per the policy exception (T-call required, insurance re-verified, driver hours confirmed). That connects the exception record to the specific load event.

A reminder in the monitoring queue to re-run the screen at the next tender for this carrier. A policy exception approved today isn't a standing approval. It's approval for this load, this date.

The stack of paper isn't protection from every lawsuit. But it's the difference between a lawyer who can find your reasoning and one who can only find your silence.

— Mason Lavallet

Founder, DOTScreener.com