Flexport links AI productivity claims to customs controls and freight growth
Flexport founder Ryan Petersen said freight volumes were up 40% year over year while headcount grew 5%, attributing the gap to AI used in customs, routing and engineering.
The workflow includes a tariff-classification tool that proposes 10-digit codes and links CBP rulings, while customer-specific delivery rules are converted into code after expert review. Licensed brokers still approve customs filings because U.S. law reserves customs business to qualified brokers.
Flexport reported a 0.2% internal filing error rate and cited a 20% average error rate from its audits of competing broker entries, while acknowledging that outdated CBP rulings can mislead models.
Why it matters: The Flexport productivity claim matters because forwarding growth is being tested against customs accuracy, broker capacity and compliance exposure rather than headcount alone.
Route product descriptions, prior rulings, customer rules and shipment milestones through a broker-reviewed classification queue; track correction rate, clearance time, audit findings and volume per operator.
Suggested executive takeaway: Flexport should publish lane-level evidence separating AI productivity from volume mix, tariff conditions and broker review capacity.