The Voice on the Phone Isn't Your Dispatcher Anymore: AI Voice Cloning and the New Wave of Freight Fraud

A dispatcher picks up the phone. It's the ops manager's voice, stressed, asking for a rate confirmation to be re-sent to a "corrected" bank account, or for a load to be re-assigned to a different carrier because the original truck "broke down." The voice sounds right. The urgency sounds right. Nothing about the call feels off.

Except the ops manager never made that call. A cloned version of their voice did.

Why this is suddenly a trucking problem, not just a bank problem

For years, freight fraud ran on written deception: spoofed emails, lookalike domains, doctored rate confirmations. Carriers have gotten better at training staff to spot those. So the fraud has moved to a channel almost nobody has been trained to distrust: the phone call.

According to NMFTA's 2026 Transportation Industry Cybersecurity Trends Report, criminal groups targeting brokers, carriers and shippers have started layering AI-generated emails, deepfake voice calls, and spoofed dispatch updates on top of traditional business email compromise, using the calls to redirect shipments or push through fraudulent detention and lumper fee payments. NMFTA's reporting also notes that cyber intrusion increasingly precedes or directly enables physical cargo theft, which is exactly why a fake phone call is no longer just an IT problem — it's a freight-on-the-ground problem.

The technology behind it has gotten uncomfortably good. Security researchers now put the amount of audio needed to clone a convincing voice at a few seconds — pulled from a voicemail greeting, a podcast, a conference call, or a video posted online. As NMFTA cybersecurity engineer Ben Wilkens has written, the skill level and budget needed to produce a convincing fake of a trusted colleague's voice or face have dropped so far that even an average attacker can pull it off quickly and cheaply. And the World Economic Forum's Global Cybersecurity Outlook 2026 found that 87% of organizations named AI-related vulnerabilities their fastest-growing cyber risk of 2025 — not because of nation-state hackers, but because attackers are moving into the gap left by companies adopting AI-adjacent tools faster than they can govern them.

At the same time, the FBI's Internet Crime Complaint Center added a dedicated AI-fraud section to its most recent annual report for the first time, citing over 22,000 complaints involving AI-generated voice or video in a single year, with reported losses approaching $893 million. That's not a trucking-specific number, but it shows the scale of the tooling now available to whoever decides to point it at a dispatch desk instead of a bank.

What this looks like inside a carrier or brokerage

The pattern NMFTA and other researchers are seeing follows a familiar shape, just with a more convincing voice attached:

• ★The "urgent reroute" call. A voice that sounds like a known dispatcher or broker calls in asking for a load to be released to a different carrier or driver, citing a breakdown or schedule conflict — the voice-call version of the broker-impersonation email scams already hitting carriers.
• ★The "corrected payment details" call. A cloned executive or accounting voice calls a bookkeeper or factoring desk to change ACH routing details on an invoice — the same playbook behind ACH-redirect and BEC fraud, just delivered by phone.
• ★The fraudulent detention or lumper fee call. A voice claiming to be a shipper or receiver contact pushes a dispatcher to approve an unusual detention or lumper payment on the spot, exploiting the fact that these are often processed quickly and with less scrutiny than a standard invoice.
• ★The internal "leadership" call. A voice cloned from an owner or ops director calls a member of staff directly, bypassing normal approval chains by invoking urgency and authority — the same pressure tactic behind dispatcher-targeted phishing emails, just harder to second-guess.

Why trucking is a soft target for this

Dispatch desks are built for speed. Loads move on tight windows, detention clocks are running, and a dispatcher who hesitates on a legitimate urgent call costs the company money. That culture of "act now, verify later" is exactly what voice-cloning fraud is designed to exploit. Add in a workforce that's used to unfamiliar numbers calling about loads, brokers, and carriers they've never worked with before, and you have an environment where a confident, familiar-sounding voice rarely gets questioned.

It also compounds the risks carriers are already dealing with. A cloned voice used to talk a dispatcher into a fraudulent load reassignment plugs directly into the kind of cyber-enabled cargo theft already costing the industry hundreds of millions of dollars a year, and a cloned voice used against accounting plugs into the same payment-fraud pipeline carriers are already fighting on the email side.

What carriers can do about it, starting Monday

Voice cloning defeats "does this sound right" as a verification method. It doesn't defeat a process.

The bottom line

The written scams carriers have spent the last two years learning to spot are being joined by a spoken version that's harder to distrust on instinct. The fix isn't teaching people to recognize a fake voice — most people can't, and even trained security researchers struggle. The fix is building a callback-and-confirm habit into every high-risk request, so the voice on the other end of the line never gets the final word.