C.H. Robinson Did the Vetting. The Jury Awarded $604 Million Anyway.

In March 2021, a tractor-trailer operated by Lupus Superior, LLC — a carrier based in Grand Prairie, Texas — struck stopped traffic on Interstate 20 in Warren County, Mississippi, triggering a multi-vehicle fire. Three people died: Jennifer Lipe, Benjamin Brewer, and Rhoderick Coleman. Two others survived with injuries. C.H. Robinson had brokered the load.

The case was tried in Dallas County, Texas (Case No. CC-22-07326-D, Judge Dianne Jones presiding), not in Mississippi — venue followed the carrier's home state. If you've seen "Mississippi court" in a forum thread, that's the crash location, not the courthouse.

After a roughly two-week trial and three days of deliberation, the jury returned $604,025,000 on July 23, 2026. (C.H. Robinson disclosed the verdict publicly the next day, which is why some coverage dates it July 24.) The award breaks out by family as shown below.

All of it is compensatory. There is no punitive component — C.H. Robinson's own securities disclosure describes an advisory verdict of compensatory damages of $604 million. That surprises people, and it matters: punitive damages are the part that gets reduced by statutory caps in a lot of jurisdictions. Compensatory awards for three deaths don't have that same ceiling.

Fault was apportioned: driver 45%, Lupus Superior 32%, C.H. Robinson 23%. The carrier was found liable too. And one detail worth holding onto: the jury was asked whether C.H. Robinson was negligent in undertaking the responsibilities of a motor carrier, and answered no. This was not a finding that the broker was secretly operating as a trucking company.

C.H. Robinson's chief legal officer, Dorothy Capers, said the company "strongly disagree[s] with the verdict" and will "immediately appeal," noting the carrier "had safely delivered nearly 270 loads for our customers and held a Satisfactory FMCSA rating when we selected it," and that the rating remained Satisfactory after a federal review of the accident.

The verdict is not final. "Advisory" is the company's own word for it in that disclosure; post-trial motions come first, and the trial court has not yet entered judgment. Any appeal follows that. We're not going to speculate about how it turns out.

If the verdict had stopped at 23%, this would be a large but structurally ordinary negligent-selection case. It didn't stop there.

The jury also found that the driver was a "borrowed employee" of C.H. Robinson — that he was, in the charge's language, operating the vehicle in furtherance of a mission for the broker's benefit and subject to the broker's control as to the details of that mission.

That finding does something different from a negligence finding. Negligence asks whether you were careless. Vicarious liability asks whether the person who was careless was, legally speaking, yours. Answer yes, and their share becomes your share. The driver's 45% lands on the broker on top of its own 23% — roughly 68% of the total.

So there are two separate questions in this case, and they have different answers. Did the broker exercise reasonable care in picking this carrier? The jury said not entirely — 23%. Was the driver effectively working under the broker's control? The jury said yes — and that's the finding that turned a nine-figure exposure into a much larger one.

Legal commentators generally read the borrowed-employee finding as the most likely target on appeal, since it turns on the sufficiency of the evidence about control. We'll leave it there.

This is the part small brokers keep asking about, and it's the most useful thing in the case.

Lupus Superior's Satisfactory rating had been in place since 2014, reaffirmed in 2021 and again as recently as April 2026. By the ordinary standard — check the rating, check authority, check insurance — the carrier passed.

Plaintiffs argued the rating wasn't the whole picture. Their safety expert, Prof. Thomas Corsi, testified that during every month of the year preceding the crash (March 2020 through March 2021), Lupus Superior exceeded FMCSA's intervention thresholds in Unsafe Driving and Hours-of-Service Compliance — a fact plaintiffs said C.H. Robinson knew or should have known. Plaintiffs also alleged the driver had falsified logs, and contended that the driver notified both the carrier and the broker that he was too sick to keep driving, and that the load continued anyway. C.H. Robinson disputes the characterization of its conduct.

Two things follow from this, and they pull in opposite directions.

A safety rating is a status, not a pattern. It's a point-in-time federal determination that changes only after a compliance review. The roadside data underneath it moves continuously, and it can move a long way while the rating sits still. If your process ends at "Satisfactory," you're reading the slowest-moving field on the page. (We wrote about the mechanics of that lag separately — the inspection reaches the record fast; the score doesn't.)

And some of what the jury saw, you can't see. The Transportation Intermediaries Association has pointed out that the carrier was reportedly above the intervention threshold in two BASIC categories, which would place it on an internal FMCSA list of carriers prioritized for investigation — a list that is not public and that no broker, and no vendor, can pull. This is the crux of TIA's petition to FMCSA: brokers are being held to a standard that requires data the agency doesn't publish.

We want to be exact about this, because it's easy to oversell: BASIC percentiles and roadside inspection history are public and checkable. The internal high-risk designation is not. A tool can surface the first. Nothing surfaces the second.

For years brokers had a strong federal shield. The FAAAA (49 U.S.C. § 14501(c)(1)) blocks state laws "related to a price, route, or service" of a broker, and brokers argued a negligent-selection claim was exactly that. Courts split — the 9th and 6th Circuits let such claims proceed under the statute's safety exception; the 7th and 11th held them preempted.

That split closed on May 14, 2026, when the Supreme Court decided Montgomery v. Caribe Transport II, LLC 9–0. Justice Barrett's opinion held that a state-law negligent-selection claim against a broker is not preempted, because requiring a broker to use ordinary care in choosing a carrier concerns motor vehicle safety. (Montgomery, it's worth noting, was also a C.H. Robinson case — the carrier there held a Conditional rating, and the Seventh Circuit had ruled for the broker before the Supreme Court reversed. The same defendant opened the courtroom door and took the first nuclear verdict through it.) We've covered what the Montgomery Court actually held and what it means for a small brokerage in detail.

Two things about Montgomery are widely misread. It did not create a federal vetting standard — "reasonable care" is still defined state by state, and there is no checklist that federal law blesses. And it did not address vicarious liability at all; Justice Kavanaugh's concurrence explicitly cautioned that the decision should not be read to mean brokers will routinely be subject to state tort liability.

Which is why Lipe landed the way it did. Montgomery unlocked the courtroom door for direct negligence claims. The borrowed-employee theory in Lipe is a separate, older doctrine that Montgomery never touched — and it's the one that produced the number.

Here is the honest version, and we'd rather lose a signup than blur it.

C.H. Robinson vetted this carrier and lost anyway. They had the rating. They had the history. They have a compliance department, outside counsel, and a documented process that a solo broker will never match. None of it prevented the verdict.

Documentation does not prevent liability. It does not prevent being sued. It does not make you safe.

What it does is narrower and still worth having: when your defense counsel asks what you knew and when you decided, you can answer with records instead of recollection. That's the whole claim. Most small brokers, honestly assessed, cannot answer that question today — the check happened in a browser tab that closed, the decision lived in someone's head, and the file gets assembled after the claim rather than before it.

Records built at the time of the decision carry weight that reconstructions don't. Records assembled after a demand letter carry the least.

Concrete, in rough order of how much they matter relative to effort:

Draylo checks a USDOT number against live FMCSA data — authority, insurance filings, out-of-service status, roadside inspections — re-checks every carrier in your book daily, and alerts you when something changes.

The part relevant to this post is the record it leaves behind. Every check and every approve/deny decision is written to a tamper-evident audit trail: content-hashed, anchored daily, independently timestamped, exportable as a PDF. Your gate rules function as the written standard, applied the same way every time, with the exceptions logged.

That means if you're ever asked what the record showed on the day you tendered a load, the answer is a document rather than a memory.

It does not mean you won't be sued, and it does not mean you'll win. C.H. Robinson is the proof of that. It means you can show your work.