Payment diversion fraud: the controls that stop it
Criminals redirect client money by changing bank details in an email. The payment and IT controls that stop it, and where responsibility sits.
A criminal does not need to break into your case management system to steal completion funds. They need one person to act on an email that says the bank details have changed.
The National Crime Agency calls this payment diversion fraud. It is also known as mandate fraud, or business email compromise, which simply means a fraud run through a hijacked or faked email account. The NCA describes it as criminals impersonating others, creating or amending invoices, and diverting payments to bank accounts under their own control.
For a law firm, the sums are large, the window is narrow and the money is not yours. When the NCA and the Law Society launched a joint campaign for solicitors and conveyancers on 30 September 2025, the NCA said Action Fraud figures showed 143 reports of payment diversion fraud during property transactions between April 2024 and March 2025, with victims losing an average of £82,000.
This guide covers what the fraud looks like in practice, the payment checks the NCA asks for, what IT controls genuinely do and do not do, and where responsibility sits. If you want the wider picture of how we work with firms like yours, see our page on IT support for law firms.
One thing to be clear about before we start. This is a description of what the National Crime Agency, the Solicitors Regulation Authority, the National Cyber Security Centre, Pay.UK and Microsoft publish on their own pages, and of the IT work that sits behind it. It is not legal or regulatory advice. No IT provider, including us, can make your firm compliant with the SRA’s rules. Those obligations are yours and they stay yours.
How the fraud actually works
There are three common shapes, and they are often combined.
The mailbox is read. Someone signs in to a mailbox they should not have access to, usually with a password taken in an earlier phishing email. They do not send anything at first. They read. They learn the completion date, the amounts, the names, the house style. Then they set up an inbox rule, a small automatic instruction inside the mailbox, so that messages from the other side are quietly copied out or moved out of sight. The SRA describes exactly this, noting that attacks on firms included fraudsters monitoring and intercepting emails, for example by using email forwarders during a transaction.
A lookalike domain is registered. The email comes from a domain that reads like the real one at a glance, with a letter changed or added. This matters technically, and we come back to it below: Microsoft’s own documentation says impersonation can pass the standard email authentication checks if the attacker created a lookalike domain and published valid records for it. In other words, the checks that stop someone forging your domain do not stop someone registering a near-copy of it.
The pressure is applied. The NCA’s four red flags are worth pinning above a desk. Have you been asked to urgently process a payment that is large or unusual? Have you been asked to change the bank details of an existing supplier, or to set up a new one? Is the language inconsistent with the genuine sender? Are there spelling mistakes in the email or the address?
The SRA publishes a case study that shows how thin the margin is. A firm acting on a property sale received, on the day of exchange, a modified email purporting to come from its own client, changing instructions and asking for half the completion funds to go to a different account at the same bank. The firm’s policy said staff should get verbal confirmation of any change to bank details received by email, and they called the client. The client was busy, did not have the details to hand, and told them to get on with it. The employee actioned the payment without ever receiving confirmation.
That case study is the SRA’s, not ours. We have no client case studies to publish on this.
The payment checks the NCA asks for
The campaign with the Law Society reduces to three steps, in the NCA’s own words:
- CHECK by calling before you transfer money, as emails can be intercepted or diverted.
- TEST the account is genuine by sending a small sum to the account details provided and ensure it has been received correctly.
- NEVER transfer money until you are satisfied the details are correct.
The NCA and Law Society info sheet adds the practical detail that makes those steps work. Get bank details directly, in person or on the phone, at the start of the matter rather than at the end. Question the authenticity of any email or call that says bank details have changed, because, as the info sheet puts it, law firms rarely change their bank details. If you need to confirm something with another firm, call a senior person on their published number, not the number given in the email asking for the money. Tell your clients to check details directly with you before sending funds, and ask them not to read transaction email on public or unprotected Wi-Fi.
The SRA puts the client-facing half of that even more simply. Its advice to firms is to tell your clients that you would never change your bank details by email, and its checklist asks whether you warn them of that.
Confirmation of Payee, and the gap in it
Confirmation of Payee is the check that shows the account holder’s name when you set up a payment. Pay.UK, which owns and runs it, describes it as an account name checking service designed to reduce misdirected payments. It returns one of four answers: a match, a close match where you are shown the real name to compare, a flat no, or unavailable, which covers a timeout, an opt-out or an account that does not exist. For a business account you need the registered or a trading name. Pay.UK says more than 140 banks, building societies and payment firms offer it, and that where an organisation does not, you get an “unable to be checked” message.
It is useful and it is not a solution. Two things to understand about it.
First, Pay.UK says plainly that it does not automatically guarantee a fraudulent payment will be detected, or that a victim will be reimbursed, and that it is not a silver bullet.
Second, and this is the part that catches people, Pay.UK says the service checks new payees and amendments to existing payee details, and that payments to existing recipients and standing orders may not be checked, although individual banks may choose to check them. So the check is most likely to fire at exactly the moment the fraud happens, which is good. But a name match only tells you that the name on the account matches the name you typed. It does not tell you that the account belongs to the firm you believe you are paying. If the criminal has opened an account in a name close enough to pass, a match is not reassurance.
What IT controls actually do
Five things, and it is worth being precise about each, because the honest answer to “will this stop it” is different in every case.
Keep the mailbox from being taken over. Almost every version of this fraud starts with a stolen password. Multi-factor authentication, which means a second proof of identity on top of the password, is the control that makes a stolen password much less useful. Not all second factors are equal, and text message codes are the weakest of the common ones. We cover where Microsoft is taking sign-in, and what to do about it, in our guide on Microsoft passkeys and the retirement of SMS codes.
Block automatic forwarding out of the firm, then check nobody has set one up. This is the single clearest case where a professional checklist question lands on a specific setting. The SRA asks firms: have you checked whether your email account has been compromised with forwarder rules? In Microsoft 365, that control sits in the outbound spam filter policy, and there are three possible values. Microsoft’s documentation is worth quoting because the default is the trap: the default value is “Automatic - System-controlled”, and Microsoft says that because the behaviour can differ between organisations, you should configure it explicitly to on or off instead of leaving it on automatic. Depending on when your tenant was set up and how it has been used, automatic may mean forwarding is allowed.
Set to off, Microsoft says, it disables any inbox rules created by users and any mailbox forwarding set by administrators that redirect messages to external addresses, and the sender gets a bounce message reading “5.7.520 Access denied, Your organization does not allow external forwarding”. Forwarding between colleagues inside the firm is not affected. There is also a report, the auto forwarded messages report, that shows which users are forwarding mail outside the organisation, which is how you answer the SRA’s question with evidence rather than a shrug.
Make your own domain hard to forge. The NCSC’s anti-spoofing guidance sets out three records you publish in your domain’s settings: SPF, which lists the servers allowed to send mail as you; DKIM, which adds a signature a receiving server can verify; and DMARC, which tells receiving servers what to do when the first two fail, and reports back on what is being sent in your name. The NCSC’s recommended route is to start DMARC on a monitoring-only setting while SPF and DKIM are made reliable, then tighten it. The SRA’s own cyber security page points firms at DMARC for this reason. What this stops is somebody sending mail that claims to come from your domain. What it does nothing about is the lookalike domain, which is the next item.
Catch lookalike senders, and know what you are paying for. Impersonation protection is the feature that watches for senders whose name or domain is a near-copy of someone important. Two things about it are routinely misunderstood. It is not part of the protection that comes with every Microsoft 365 mailbox: Microsoft’s documentation states that anti-phishing protection for all cloud mailboxes does not include impersonation protection, and that it requires Microsoft Defender for Office 365. And it does nothing until somebody fills it in. Microsoft says that by default, no sender email addresses and no sender domains are configured for impersonation protection, in either the default policy or a custom one. Someone has to list the people worth protecting, your partners and your accounts team, and the domains worth protecting, your own and those of the firms you deal with most. Microsoft’s limits are 350 users and 50 custom domains per policy, which is ample for a firm of any size we work with.
Two caveats from the same documentation, so there are no surprises. Where mailbox intelligence and intelligence for impersonation protection are both switched on, user and domain impersonation protection does not apply when the sender and recipient have emailed each other before. And the first contact safety tip, the banner telling a reader they do not often get email from this address, is something Microsoft explicitly recommends turning on. For conveyancing and accounts inboxes, it earns its place.
Notice a compromise in hours, not at completion. All of the above is prevention. The remaining question is how long an account takeover would go unseen. A security operations centre watching sign-ins, mailbox rule changes and unusual access around the clock is how that gets caught while it still matters. That is our managed detection and response service, which is a standalone service and also included in Lanmark Total.
What IT controls cannot do
None of it stops a person acting on a plausible email. The SRA’s case study had the verbal confirmation policy, made the call, and still paid. Controls reduce the number of convincing emails that reach people, remove the quiet forwarding that lets a criminal watch a transaction, and shorten the time a compromise goes unnoticed. The decision to release money is still a human one taken under time pressure, usually by the person with the least authority to say no. That is why the NCA puts the phone call first and why training the people who touch payments is not optional.
It is also why “we have Microsoft 365, so we are covered” is the sentence we hear most often and agree with least. Several of the settings above are off, or ambiguous, until someone deliberately changes them.
Where responsibility sits
The SRA’s position, on its own pages, is that its Accounts Rules require a firm to immediately pay in or replace money improperly withdrawn from the client account, and that firms should report the loss to the SRA. Those duties belong to the firm. They cannot be contracted out to an IT provider, and no product or service makes a firm compliant with anything.
What an IT provider can honestly take on is narrower and still worth having. We can tell you what is configured on your systems today and show you the evidence. We can change the settings and keep them changed. We can watch for the signs of a mailbox compromise and tell you quickly. We can train the people who handle payments and test them. We hold Cyber Essentials, the government-backed certification covering five basic technical controls, and we have recently renewed it. We will not tell you what the SRA requires of you, because that is not ours to say.
A checklist to take to your IT provider
Each of these has a yes or no answer with evidence behind it. If the answer to any of them is “I will find out”, that is useful information about your provider.
- Is automatic external forwarding in Microsoft 365 set explicitly to on or off, rather than left on the system-controlled default?
- Can we see the auto forwarded messages report, and does anybody in the firm currently forward mail outside it?
- Does every account have multi-factor authentication, and what are the exceptions?
- Do we publish SPF, DKIM and DMARC records, and what is the DMARC policy set to?
- Do we have Microsoft Defender for Office 365, and if so, which people and which domains are actually listed for impersonation protection?
- Is the first contact safety tip turned on for the conveyancing and accounts inboxes?
- Who gets told, and how quickly, if a mailbox rule is created that sends mail outside the firm?
- If a mailbox were compromised this morning, how would we find out, and by when?
- Which of the people who handle payments have had phishing training in the past 12 months, and what did the last test show?
Tell us which systems you use for case management, accounts and client money and we will confirm what we can support.
If you are not a law firm
The same fraud, the same controls, different paperwork. Any business that pays suppliers by bank transfer is exposed to the version that arrives as a supplier’s change of bank details.
The NCA publishes a case study worth reading if you run a finance function. A City of London firm found 30 fake invoices that appeared to have been approved for payment, going to 21 different bank accounts and totalling £670,000. Some payments were stopped and the loss was about £340,000. The emails appeared to come from the PA to the chief executive and had in fact been sent through a spoofing website. What gave it away afterwards was the timing: the invoices had been processed within minutes of arriving, where the firm’s normal practice was to process similar invoices at least a day later.
The national fraud reporting hub adds two sensible process controls for this version. Restrict changes to payment details to designated people rather than anyone who opens the post. And if you are hit, hang up on anyone who later rings offering to recover the money, because that is a second fraud aimed at the same victim.
If it has already happened
Act in this order. Tell your bank immediately and ask them to contact the receiving bank, because the chance of recovering funds falls away quickly. Report it through the national fraud reporting hub at reportfraud.police.uk or by calling 0300 123 2040. Tell your IT provider, because if a mailbox was compromised, other matters and other clients may be exposed. If a mailbox was involved, our guide on what to do in the first hour after a suspected breach sets out the technical steps, including the one most people get wrong, which is wiping or rebuilding the machine before anyone has looked at it.
Then there is the regulatory side, which is yours: the SRA says firms should report the loss of client money to it, and if client personal data was exposed there are separate duties. We are not the people to advise you on either.
Where Lanmark fits
We configure and check the settings above, we show you the evidence rather than asserting it, we run the security operations centre that watches for a compromised account, and we train the people who touch payments. We hold Cyber Essentials and have recently renewed it.
We do not advise on the SRA Accounts Rules, we do not make your firm compliant, and we will tell you plainly if your current arrangements are already sound and you do not need to change anything. If you want someone to go through the nine questions above with you, book a free IT review.
Sources
Every fact on this page was checked against these pages on 28 October 2026.
- National Crime Agency, “NCA and The Law Society launch campaign to protect against Payment Diversion Fraud in property sales”, 30 September 2025. The campaign, the three steps, the Action Fraud figures for April 2024 to March 2025.
- National Crime Agency and The Law Society, payment diversion fraud info sheet. Getting bank details at the start, published numbers, client advice.
- National Crime Agency, “Payment Diversion Fraud? What you need to know, and how to protect yourself”. The definition, the four red flags, the City of London case study.
- Solicitors Regulation Authority, cyber security advice for solicitors. Email modification fraud, forwarder rules, the conveyancing case study, telling clients you never change bank details by email, the Accounts Rules position and reporting a loss. This page dates from the SRA’s thematic work of 2019 and 2020, so we have used its advice and not its figures.
- National Cyber Security Centre, “Email security and anti-spoofing”, section 3. SPF, DKIM and DMARC.
- Pay.UK, Confirmation of Payee frequently asked questions. What the service checks, the four outcomes, coverage, and what Pay.UK says it does not guarantee.
- Microsoft, “Control automatic external email forwarding from cloud mailboxes”. The three settings, the ambiguous default, what switching it off does, and the auto forwarded messages report.
- Microsoft, “Anti-phishing policies in cloud organizations”. Impersonation protection, what is off by default, the limits, and lookalike domains passing authentication checks.
- National fraud reporting hub, reportfraud.police.uk, mandate fraud. Reporting route, designated staff for payment changes, and recovery scams.
Common questions
What is payment diversion fraud?
The National Crime Agency describes payment diversion fraud, also known as business email compromise or mandate fraud, as criminals impersonating others, creating or amending invoices, and diverting payments to bank accounts under their own control. In a law firm it usually appears as an email saying bank details have changed shortly before completion funds or a deposit are due to move.
How much do firms and clients lose?
The National Crime Agency, announcing a joint campaign with the Law Society on 30 September 2025, said Action Fraud figures showed 143 reports of payment diversion fraud during property transactions between April 2024 and March 2025, with victims losing an average of £82,000. That is one 12-month period and the only figure for it we can source, so treat it as the scale of the problem rather than a current trend.
Does Confirmation of Payee stop it?
Not on its own. Pay.UK, which runs the service, says it is an account name checking service, that it does not automatically guarantee a fraudulent payment will be detected or that a victim will be reimbursed, and that it is not a silver bullet. It also says it checks new payees and amendments to existing payee details, and that payments to existing recipients and standing orders may not be checked. A name match tells you the name on the account matches what you typed, not that the account belongs to the firm you think you are paying.
Can IT settings prevent it?
They can remove several of the steps a criminal relies on, and they can shorten the time a compromised mailbox goes unnoticed. They cannot stop a person acting on a plausible email. The Solicitors Regulation Authority publishes a case study in which a firm's policy required verbal confirmation of any change to bank details, staff called the client, could not get confirmation, and sent the money anyway. Controls reduce the number of convincing emails that reach people. The decision to release money stays human, which is why the National Crime Agency puts the phone call first.
Does using an IT provider make our firm compliant?
No. No IT provider, including us, can make your firm compliant with the SRA's rules, and we do not give legal or regulatory advice. We can tell you what is configured on your systems, change the settings, show you the evidence and watch for signs of compromise. Reporting a loss of client money, and the obligations under the SRA Accounts Rules, stay with the firm.
Where to go next
About this guide
Written by Lanmark, Lanmark Limited, 36 Basepoint, Victoria Road, Dartford, Kent DA1 5FS. Published .
Book a free IT review.
In about half a day we look at what you have, what it costs and where the risks are, and give you a red, amber and green (RAG) report. We tell you plainly if you do not need to change anything.
Book a free IT review