
For nearly two years, a single fraud campaign has been hopping from state to state. First dressed as a toll-road violation, it is now dressed as a notice from the Department of Motor Vehicles.
The newest version warns drivers that their license is about to be suspended unless an outstanding "fine" is paid by the end of the day. There is no fine. There is no DMV behind the text. And the message is more convincing than it used to be: the misspellings and broken grammar that consumers were taught to look for are gone.
Artificial intelligence is writing the new ones.
This case file examines the Oregon DMV warning, explains why this scam represents a dangerous evolution in AI-powered fraud, and provides concrete protections for every driver with a smartphone.
In a May 9, 2026, report by Ginnie Sandoval of the Statesman Journal / USA TODAY NETWORK, the Oregon Department of Transportation laid out an updated warning about a nationwide scam campaign. The warning built on an official ODOT release issued May 8, 2026, and supplemented existing fraud-prevention resources on the agency's website.
"If you've received text or email messages claiming to be from Oregon DMV saying you owe money, you're not alone," the agency said in its statement. "This nationwide scam, which began as a toll phishing scam, has been targeting drivers in several states for nearly two years."
The mechanics are consistent across the campaign:
The campaign began as a toll-road phishing scam in 2024, sending messages to drivers in states with electronic tolling systems (Massachusetts, New York, Illinois, California, Florida). The message claimed the driver had an unpaid toll balance and faced escalating fines or legal action.
When that campaign started to lose effectiveness, as state transportation agencies issued warnings and the public became aware, the scammers pivoted. The same infrastructure, the same phone numbers, the same fake payment portals were repurposed for a new pretext: unpaid DMV fees and license suspension threats.
The Oregon DMV's most striking observation concerns the quality of the scam messages themselves.
"In the past, scam messages were often filled with spelling and grammar mistakes. Many scammers are now using artificial intelligence to create more realistic, professional-sounding messages."
This is the same shift AuthentiLens has documented across multiple fraud categories in recent weeks. The grammatical errors that once served as a reliable tell for scam communications have disappeared. AI language models can now generate flawless prose in any language, at any length, with any desired tone.
ODOT was unambiguous about its own behavior. The agency's statement made clear:
Real Oregon DMV transactions occur only through:
The agency directed drivers uncertain about a message to call Oregon DMV directly at 503-945-5000 or ODOT's customer service line at 1-888-Ask-ODOT.
The Oregon DMV warning is not about a new scam. It is about an evolved scam, one that has been adapting for nearly two years.
The campaign began in 2024 as a toll-road violation scam. Drivers in states with electronic tolling systems, including Massachusetts (E-ZPass), New York (Tolls by Mail), Illinois (I-PASS), California (FasTrak), and Florida (SunPass), received texts claiming they had unpaid toll balances.
The messages typically read something like:
"E-ZPass Notice: You have an outstanding toll balance of $8.75. Failure to pay by [date] will result in a $75 late fee and potential suspension of your registration. Pay now at [fake URL]."
The scam was effective because:
State transportation agencies quickly issued warnings. The FBI's IC3 logged thousands of complaints. But the scammers did not stop. They adapted.
When the toll-road pretext became too widely recognized, the scammers pivoted to a new cover story: unpaid DMV fees and license suspension threats.
The new messages read something like:
"Oregon DMV Notice: Your driver's license is scheduled for suspension due to an unpaid administrative fee of $12.50. Payment must be received by 5:00 PM today to avoid suspension. Update your account at [fake URL]."
The pivot was strategic:
The Oregon DMV warning notes that the same scam network may pivot again. "When state DMV impersonation is no longer profitable, they will find another agency to impersonate," a DMV spokesperson told the Statesman Journal. "The infrastructure is the same. The text changes."
Security researchers tracking the campaign expect the scammers to continue pivoting. Potential future pretexts include:
The cost of each pivot is near zero. The scammers simply change the text of the message, update the fake landing page, and purchase a new lookalike domain. AI can generate the new copy in seconds.
The Oregon DMV's observation about AI-written messages is not speculative. Researchers have documented the use of large language models in phishing campaigns since 2024. What has changed is the scale and quality.
For years, consumer protection advocates taught people to spot scam messages by looking for:
These tells worked because many scammers were operating in a second language. They were not native English speakers. Their messages contained errors that native speakers would not make.
AI has eliminated that tell. Large language models can generate flawless prose in any language. The scam text you receive may be grammatically perfect, stylistically consistent, and contextually appropriate. It may still be a scam.
The process is simple and cheap.
Step 1: Choose a pretext. The scammer decides on a cover story, such as unpaid tolls, DMV fees, court fines, or utility bills.
Step 2: Generate a prompt. The scammer writes a prompt for an AI language model: "Write a text message claiming to be from the Oregon DMV. The message should say the driver has an unpaid fee of $12.50. It should threaten license suspension if not paid today. It should include a link to oregon-dmv-payment.com. Make it sound urgent and official."
Step 3: Generate the message. The AI produces a message in seconds. The scammer can request variations — shorter, longer, more urgent, more formal — until they are satisfied.
Step 4: Mass distribute. The scammer sends the AI-generated message to thousands of phone numbers simultaneously, using automated texting platforms.
Step 5: Iterate. If one version stops performing, the scammer generates a new one. The cost is negligible. The time investment is minutes.
AI-generated scam texts are harder to detect for three reasons:
The Oregon DMV scam is not yet a flood. According to security researchers who track smishing (SMS phishing) campaigns, AI-written messages still represent a minority of scam texts, perhaps 1 to 5% of the total.
But that minority is growing. And because AI-generated messages are harder to detect, they are more likely to reach victims. A poorly written scam text might be deleted immediately. A professional-sounding AI message might be read, considered, and acted upon.
The Oregon DMV's decision to issue a public warning suggests that the agency has seen enough of these messages to consider them a significant threat.
The Oregon DMV warning is part of a national pattern. Similar warnings have been issued by DMVs and transportation agencies across the country.
Massachusetts RMV. In early 2025, the Massachusetts Registry of Motor Vehicles warned residents about text scams claiming unpaid E-ZPass tolls. The messages threatened license suspension and directed victims to fake payment portals. The Massachusetts RMV noted that the texts were "increasingly sophisticated" and "difficult to distinguish from legitimate agency communications."
New York DMV. In mid-2025, the New York Department of Motor Vehicles issued a similar warning about texts claiming unpaid "Tolls by Mail" balances. The agency noted that the scam had "evolved" from earlier iterations and was "using more convincing language."
Illinois Secretary of State. In late 2025, the Illinois Secretary of State's office, which oversees driver's licenses and vehicle registrations, warned residents about texts claiming unpaid "administrative fees" that would lead to license suspension. The office noted that the messages "appear to be written by native English speakers" and "lack the grammatical errors that often indicate a scam."
California DMV. In early 2026, the California Department of Motor Vehicles warned about texts claiming unpaid "vehicle registration fees" that would result in suspension. The agency noted that the scam was "nationwide" and that "similar messages have been reported in at least a dozen states."
Every state warning shares a common observation: the scam messages are getting better. The grammar is correct. The formatting is professional. The urgency is calibrated. The telltale signs that consumers were taught to spot are disappearing.
The Oregon DMV's explicit mention of AI as the cause is notable. Most state agencies have been reluctant to name AI specifically, preferring to warn about "sophisticated" scams. Oregon's direct statement, "scammers are now using artificial intelligence to create more realistic, professional-sounding messages," is a useful public education step.
The Oregon DMV scam matters for reasons that go beyond a single text message.
The end of the "obvious scam." For decades, consumers could rely on a simple heuristic: scam messages look fake. They have typos. They have bad grammar. They come from weird email addresses. They ask for gift cards.
That heuristic is now obsolete. AI-generated scam messages look real. They have perfect grammar. They come from spoofed numbers that may match legitimate agency caller IDs. They ask for credit card payments on websites that look exactly like government portals.
The consumer who is looking for obvious signs of fraud will not find them. The scam will look legitimate. That is the danger.
The agency impersonation problem. The Oregon DMV scam exploits a specific vulnerability: most people do not know exactly how their state DMV communicates. Does the DMV send text messages? (No, in Oregon's case.) Does the DMV send email reminders? (Generally no for payment requests.) How are license suspensions communicated? (By mail, not by text.) The average driver cannot answer these questions. The scammer is counting on that ignorance.
The erosion of trust in government communications. The long-term cost of these scams may not be the money lost, though that is substantial. It may be the erosion of trust in government communications. When drivers receive a text that looks like it is from the DMV but is actually a scam, they learn to distrust all DMV communications. When the real DMV sends a legitimate notice by mail, the recipient may assume it is also a scam. When state agencies need to reach residents about genuine issues, their messages may be ignored or deleted.
This is the same dynamic that has played out with IRS phone scams. The IRS does call some taxpayers (contrary to popular belief), but the prevalence of IRS-impersonation scams has made many people unwilling to answer any call claiming to be from the IRS, even legitimate ones.
This Oregon DMV case connects directly to the pattern AuthentiLens has tracked in recent weeks. The Chicago man who lost $69K to an AI-generated U.S. Marshals badge scam , the AI scammers holding fake immigration court hearings , and the FTC's $2.1 billion social-media-scam disclosure all share the same core pattern: AI tools lowering the cost of impersonation while raising the quality of the deception. Now the same shift has reached every driver's phone.
The AuthentiLens editorial team has distilled the Oregon DMV warning and the broader smishing pattern into six concrete protections for every driver with a smartphone.
This is the single most important rule. If you receive any text claiming to be from the Oregon DMV, or any other state DMV, demanding a fee or threatening license suspension, it is fake. There is no exception.
The same rule applies to text messages claiming to be from:
Government agencies do not use text messages to demand payment. They use mail, official letters, in-person notifications at DMV offices, and secure online portals that you access by typing the URL yourself, not by clicking a link in a text. For more details, see our guide on signs of an impersonation scam .
When you receive a suspicious text claiming to be from the DMV, your bank, or any other organization, do not just delete it. Forward it to 7726, the keypad letters for "SPAM."
This shortcode is used by all major mobile carriers (AT&T, Verizon, T-Mobile, Sprint) to collect scam text data. When you forward a scam text to 7726, the carrier analyzes the message, identifies the originating number, and adds it to their spam-blocking databases.
This single action helps:
After forwarding the message to 7726, you can delete it.
The link in a scam text leads to a fake website designed to steal your credit card information, Social Security number, or login credentials. The phone number leads to a scam call center staffed by people who will try to extract even more money from you.
If you have a genuine question about your driver's license, vehicle registration, or any other DMV matter:
For more detailed guidance, see how to tell if a text message is a scam and how to check if a link is suspicious .
Scam texts almost always include a link to a fake payment portal. The link is designed to look like a legitimate government website, but it is not.
Real Oregon DMV transactions live at:
A fake link might look like:
Look for the actual domain name, the part immediately before the .com, .gov, .org, or .net. If that domain is not oregondmv.com or oregon.gov, it is fake.
The same rule applies to links in scam texts impersonating any other agency. Look at the domain before you click.
If you clicked a link in a scam text or provided payment information, do not panic. But act quickly.
You are not expected to become a smishing detection expert. That is what AuthentiLens is for.
When you receive a suspicious text claiming to be from the DMV, the IRS, the court, or any other government agency:
We do all of this in seconds, before you click, before you call, before you send a single dollar to a scammer.
If you have received a scam text mimicking the DMV, even if you did not click or pay, you can still take action.