As of 13 August 2026, AI can only partly check if a bank message is genuine.
This still needs a person who signs their name to it.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill needednone
Who has to check ita professional
What the alternative costsEmma is an AI budgeting app across your UK accounts, but no priced alternative for authenticating bank messages is provided here.
If this goes wrong, you may click a fraudulent link, disclose security details or send money before the bank can stop it.
What to actually do
Hand it to a person
The route this page recommends
Someone with a licence or accountable authority has to sign this before it counts.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, no skill needed, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open the message without clicking its links, opening attachments or replying, and note whether it arrived by text, email, phone call or an in-app notification.
- Copy the message text and any visible sender address or web address into a new chat, removing account numbers, passwords, PINs, one-time passcodes and other private security details.
- Paste the prompt above and the cleaned message into the chatbot, then ask it to separate observable evidence from assumptions.
- Treat the result as a phishing-risk assessment, not authentication, and do not use any link, phone number or reply address supplied in the message.
- Open your bank's official app or type its website address yourself, or call the number printed on your bank card, and ask the bank's fraud team whether the message is genuine.
- If you already clicked, shared information or made a payment, tell the bank exactly what happened through its official contact route and follow its security instructions.
Prompt
Assess the bank message below for signs of phishing or fraud. Do not say that it is genuine, safe or definitely from my bank. Separate observable evidence from guesses, quote the exact wording that concerns you, and list what cannot be checked from the message alone. Check for requests for passwords, PINs, one-time passcodes, card details, urgent payments, unusual links, unexpected attachments, threats and instructions to bypass the bank's normal app or website. Tell me not to click links, open attachments, reply or share security information. End with a short safe action plan that tells me to contact my bank through its official app, the website I type in myself, or the number on my card. Do not ask for or repeat account numbers, passwords, PINs or one-time passcodes. Message and context: [PASTE THE MESSAGE AND SAY HOW YOU RECEIVED IT, WITHOUT PRIVATE SECURITY DETAILS]
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot inspect the bank's internal records to confirm whether it sent the message.
- AI cannot reliably authenticate a sender address, phone number or web domain from appearance alone.
- AI cannot know whether a link redirects to a harmful site or whether an attachment contains malware without specialist security checks.
- AI cannot reverse a payment or accept responsibility if its assessment is wrong.
- You still have to contact the bank through an official route, which is the part that can confirm the message.
What caps this at PARTLY: real time truth, stakes of error and consent and privacy.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 1 |
| Inputs | 1 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 1 |
| Total | 5 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI tell if a bank message is real?
- It can flag common phishing signs, such as urgency, requests for security details or suspicious links. It cannot confirm that the message is genuine, so contact your bank through its official app, website or the number on your card.
- What should I paste into AI to check a bank message?
- Paste the message text, how you received it and any visible sender address or web address after removing account numbers, passwords, PINs and one-time passcodes. Do not upload private banking documents or follow anything in the message while checking it.
- Can AI check a bank email address or phone number?
- It can point out obvious inconsistencies or suspicious-looking details, but it cannot authenticate an address or number from its appearance. Check with your bank using contact details from its official app, website or your bank card.
- What should I do if I think a bank message is a scam?
- Do not click, reply, open attachments, share security information or make a payment from the message. Contact your bank's fraud team through an official route, and if you have already acted, explain what happened immediately. This is not professional advice; a serious suspected scam needs your bank's fraud team.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.