AUSTRAC Referred Brokers to Police Over Mortgage Fraud. Its Own Data Says the Channel Is Under 3%

1. What AUSTRAC announced

On 19 August 2026 AUSTRAC announced the outcome of Operation Claw, a joint analysis of data from 10 major Australian banks examining suspected mortgage fraud.

AUSTRAC described the exercise as identifying “potentially hundreds of millions” of dollars in suspected fraudulent loans. Reporting of the operation put the possible total scale of Australian mortgage fraud at around $4 billion, revised upward over the course of the work from an initial estimate of roughly $1 billion — the figure Commonwealth Bank self-reported in February 2026. Lenders prevented approximately $1.5 billion in fraudulent applications during 2025.

Names of individuals and entities were provided to law enforcement and regulatory agencies including ASIC, the Australian Taxation Office and the Tax Practitioners Board. Mortgage brokers were named among the professions referred, alongside accountants, lawyers and companies.

Trade coverage of the announcement reported that the analysis drew on approximately 1,800 suspicious matter reports lodged by more than 100 lenders, in a mortgage market of around $2.5 trillion, and that the identified activity was concentrated in New South Wales property. The techniques described included false payslips, shell companies, offshore funding sources and AI-assisted applications.

AUSTRAC chief executive Brendan Thomas was direct about who the message was for: “The scale of this activity should be a wake-up call for every lender.” He also set out the regulator’s view on where the problem is best addressed: “The most effective way to stop mortgage fraud is before a loan is approved.”

Australian Banking Association chief executive Simon Birmingham pointed to the collaboration model behind it: “This work has included intelligence sharing between banks and AUSTRAC through the Fintel Alliance.”

2. The number the headlines left out

The reporting that matters most to brokers is the channel breakdown, and it did not lead any headline.

Third-party fraud listings account for fewer than 3 per cent of all fraud listings. Over the same period, first-party channel fraud — where the applicant themselves is the source of the misrepresentation — rose to 31.6 per cent in 2025, up from 26.2 per cent in 2024.

That is a more than tenfold difference in the wrong direction for the assumption most people will take from the coverage. The channel that arranges the substantial majority of Australian residential mortgages accounts for a very small minority of identified fraud listings, and the fastest-growing category is applicants acting alone.

None of that makes the referrals meaningless. If brokers have been referred, some brokers have questions to answer, and the industry does itself no favours by pretending otherwise. But the proportion matters when you are the one having a conversation with a nervous client, a referral partner or a BDM this week.

3. A referral is not a finding

Precision is worth insisting on here, because the language will get loose quickly.

AUSTRAC provided names to law enforcement and regulatory agencies. That is a referral. It means an agency has been given information to assess. It is not a charge, not a finding, not a penalty and not a determination that any particular person did anything wrong. ASIC, the ATO and the Tax Practitioners Board each have their own processes, evidentiary standards and timelines.

This distinction is not pedantry. It governs what you can responsibly say. If a client asks whether brokers are under investigation for fraud, the accurate answer is that a regulator has referred information about individuals across several professions to authorities, that the third-party channel represents under 3 per cent of identified fraud listings, and that no findings have been made public. Anything more definitive is speculation, and speculation about identified individuals is its own risk.

It is equally worth saying plainly: if a broker has facilitated fraudulent applications, referral is the appropriate outcome, and the channel benefits from that happening. A profession that arranges most of the country’s mortgages cannot simultaneously claim the volume and disclaim the responsibility.

4. What this does to your submissions

AUSTRAC’s stated position — that the most effective intervention is before approval — has a predictable consequence. Verification tightens at the point of submission, which is your desk.

Expect movement in four areas over the coming months. This is a reasoned expectation from how lenders have responded to comparable pressure, not an announced change:

• Income evidence. Greater use of payroll and ATO-sourced verification rather than document-based checks, and less tolerance for payslips that cannot be corroborated independently.
• Employer verification. More direct employer contact and closer scrutiny of employers that appear across multiple applications.
• Entity structures. Sharper questions where borrowing entities are recently registered, non-trading, or where the substantive beneficiary of the credit differs from the named borrower.
• Source of funds. More attention to deposit origin, particularly where funds arrive from offshore or through third parties.

None of that is unreasonable. All of it slows a file that is not prepared for it. The brokers who will feel this least are the ones whose documentation already anticipates the question.

5. The exposure most brokers underestimate

There is a version of this risk that has nothing to do with a broker’s own conduct: being used.

A broker who submits a fraudulent application they had no reason to doubt has still submitted a fraudulent application. The consequences — lender review, accreditation risk, aggregator scrutiny, reputational damage — do not wait for a finding about intent. And the techniques described in this operation, including AI-assisted applications and forged payslips, are specifically designed to survive a reasonable person’s inspection.

Trade reporting has previously indicated that 73 per cent of Australian brokers experienced fraud or scam attempts in the 12 months to September 2025. On that number, being targeted is not the exception.

The protection is not suspicion of clients. It is a process that creates evidence you asked. A file that shows what you checked, when, and what you saw is a file that distinguishes a broker who was deceived from a broker who did not look. That distinction is the entire difference in how an investigation treats you.

6. Your verification hardening checklist

1. Corroborate income independently of the document. Where a lender offers payroll or ATO-linked verification, use it. Where it does not, match the payslip against bank statement credits for amount, frequency and payer name.
2. Sanity-check the employer. Confirm the ABN is active and the entity has a real presence. An employer that exists only as a name on a payslip is the single most common weak point.
3. Trace the deposit. Establish where the funds came from and record it. Genuine savings, gift, sale proceeds or borrowed funds all have different implications, and “savings” without a pattern to support it is not an answer.
4. Question recently formed entities. Where a company or trust is the borrower, check registration date, trading history and who actually benefits from the credit.
5. Watch for document artefacts. Inconsistent fonts, misaligned totals, figures that do not reconcile across documents, and metadata that does not match the stated issue date.
6. Record what you noticed. Every concern you identified and resolved should appear on the file with a date. This is the record that protects you.
7. Escalate rather than absorb. If something does not resolve, refer it to your aggregator or licensee rather than deciding alone. Your obligation is to raise it, not to adjudicate it.

7. What to watch next

• Any action by ASIC, the ATO or the Tax Practitioners Board arising from the referrals, which would move this from referral to finding.
• Lender verification policy changes, particularly around payslip and employer verification, which is where the “before approval” emphasis lands.
• The banks’ push for ATO income data access, which becomes considerably more likely to succeed after an operation of this profile.
• Whether the third-party share of fraud listings moves in subsequent reporting — it is currently the channel’s strongest defence and worth tracking.
• MFAA and FBAA responses, and whether either publishes practical verification guidance for members.