Your phone buzzes on the kitchen counter. It is a text message claiming to come from your bank's fraud department. The message says a payment of £450 to an online retailer was just attempted on your account, and asks you to reply YES to confirm or NO to stop it. Alternatively, it gives a phone number to call immediately if you did not authorise the transfer.
For years, the advice was simple: look for bad spelling, broken grammar, or strange phrasing. If a text looked sloppy, you deleted it. But today, that text looks entirely legitimate. The grammar is flawless, the tone is calm and professional, and it might even address you by your full name.
Why are scam bank texts so convincing now?
On 18 September 2026, the FBI issued a public warning about a sharp rise in sophisticated impersonation scams. According to reporting by PCMag UK and coverage from outlets such as The BayNet, scam networks are now using artificial intelligence tools to polish their text messages and phone scripts.
In the past, many scam operations operated out of call centres where workers wrote messages in a second language. Typos and awkward sentences were common red flags. Today, large language models allow anyone to generate clean, professional copy in seconds. A scammer can feed basic details into an AI tool and produce a clear, authoritative message that mimics your bank or local police force down to the precise terminology used in official communications.
How do AI impersonation scams work?
The FBI warning highlights two primary forms of contact: scam bank alerts and law enforcement impersonation.
Bank impersonation usually starts with a text about a suspicious transaction. The goal is to panic you into acting before you have time to think. If you reply NO, you might get an immediate phone call from a scammer posing as a fraud officer. Because AI can synthesize natural scripts, the caller sounds polite, patient, and completely believable. They might guide you through transferring your money to a safe account to protect it.
Police impersonation works on a similar fear. You might receive a text or call claiming that your name is linked to a financial crime or an outstanding warrant. The sender demands immediate contact to resolve the issue. The professional wording generated by AI makes the claim feel serious and official, making people far more likely to comply.
Scammers also combine AI copy with personal information stolen from online data leaks. When a text includes your real name, your post code, or the last four digits of a debit card, it feels impossible to doubt.
What are the main red flags to look for?
While AI has eliminated spelling mistakes, it cannot change the fundamental mechanics of how a scam operates. You can still spot a fraudulent message if you know what to look for.
First, look for forced urgency. Every scam relies on creating panic. The message will claim your account is about to be blocked, or that a large sum of money is leaving your account right now. Official organisations give you time to verify information. Scammers demand action within minutes.
Second, examine where the message wants you to go. Scammers need you to use the contact details they provide. They will include a link to a lookalike website or a specific phone number in the body of the text. They do not want you opening your official app or looking up a number independently.
Third, watch for requests to move money or share security codes. No legitimate bank will ever ask you to transfer your money into a new account for safety reasons, nor will they ask for your one-time passcodes over the phone.
How can you verify an urgent text without an app?
When you receive an urgent message about your money or a legal matter, the single most important habit is to pause. Never tap a link inside a text message, and never call a phone number provided in the message itself.
Instead, follow a simple rule: use a contact method you already know and trust.
If the text claims to be from your bank, turn your debit card over and ring the telephone number printed on the back. You can also log into your online banking app by opening the app directly on your phone or typing the bank address into your browser yourself. Check your recent transactions or internal inbox messages. If there is a genuine problem with your account, it will appear there.
If a text claims to be from the police or a government department, look up the main public telephone number for your local station or official agency independently. Ring them directly and ask if the message is genuine.
Taking two minutes to verify an alert through a known route stops almost every impersonation scam in its tracks.
Where Lucid fits in
If you receive a suspicious message, text, or email and want a quick second opinion before responding, Lucid offers a simple way to check it. You can submit a copy of the text, a web link, or a screenshot directly into the app.
Lucid analyses the content against known scam patterns, domain records, and reporting databases, providing a clear explanation of what the evidence shows. It gives you a neutral breakdown so you can decide how to handle the message with confidence.
Sources
- FBI Public Service Announcement on AI-enhanced impersonation scams (September 2026)
- PCMag UK report on generative AI tools used in banking fraud
- The BayNet coverage of law enforcement advisories on digital impersonation