Talk of rising mortgage fraud in Australia’s lending market — coupled with the growing use of artificial intelligence in the loan process — has put brokers on alert.
Recent reports suggest mortgage fraud has become a multi-billion-dollar problem in Australia’s lending landscape. The fraud involves borrowers allegedly falsifying or altering income documents (increasingly with the help of AI) to secure loans they would not otherwise qualify for.
Rory Sercombe, owner and broker of Melbourne-based Own Home Loans, said he recently had such an experience, where an individual posed as a borrower, booked a meeting, and then claimed the meeting link wasn’t working.
"They sent an alternative link, which was malicious, designed to deploy spyware and gain unauthorised access to broker systems and sensitive client information," Sercombe told Australian Broker. "It's a sophisticated social engineering play, and exactly the kind of attack AI is making easier to execute at scale.
"Brokers need to understand that these threats are no longer abstract," he added.
However, the issue isn't new. Australia's mortgage industry has already undergone significant regulatory tightening, with brokers held to higher standards thanks to past misconduct concerns. The 2017 Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry probe led to sweeping reforms across the financial sector, including the introduction of the Best Interests Duty (BID) law, which requires brokers to act in the best interests of their clients.
But there's still a gap. Information about individuals with prior issues does not always flow across different parts of the financial system, leaving blind spots in industry-wide checks. On the borrower side, scams are also evolving, from impersonation to altered and falsified documents. Moreover, the problem isn't confined to Australia. A recent report by Equifax revealed that banks around the world are dealing with similar AI-enabled scams.
"AI isn’t the problem," said Blake Buchanan, general manager at aggregator group Specialist Finance Group (SFG). "Rather, the people who use it for nefarious reasons are."
Even so, the evolving landscape brings new risks to brokers. So how exactly can brokers protect themselves?
In a multi-part series, Australian Broker spoke with brokers and lenders across the industry to find out what they’re seeing, and how they’re responding.
Owner and broker of Melbourne-based Own Home Loans
"There's a lot of client impersonation: AI generated voice and video cloning, which means a call or video from someone who sounds and looks like your client may not actually be them. This is a real concern that will only worsen. There are also malicious meeting links and phishing. Fraudsters use urgency and last-minute disruption to pressure brokers into clicking unverified links. The rule is simple: only ever use meeting links you generated yourself from trusted platforms.
Be careful with AI-generated documents. AI has made it alarmingly easy to fabricate payslips, bank statements and other supporting documents that look entirely convincing. Brokers should be alert to inconsistencies in fonts, formatting and figures, and where there is any doubt, verify income and account details directly with the relevant employer or financial institution rather than relying solely on what a client submits.
Brokers should also not overtrust AI tools. Brokers increasingly use AI for research and decision support. But AI can hallucinate data, cite sources that don't exist and reflect biases baked into its training. It should be a starting point, not the final word. Brokers need to treat anything AI-touched, whether it's coming from clients or being used by them, with an extra layer of scrutiny."
General manager at aggregator group Specialist Finance Group (SFG)
"The very first thing a broker should do is ensure that whichever AI tool they are using, that they immediately go into settings then data controls and switch off data sharing or ‘improve the model for everyone.’ Do it right now. And if brokers are using AI that is independent of their aggregator systems, the safest ones to use are Open AI, Microsoft, Gemini and perhaps Bedrock, all with data sharing switched off. Anything else should be used with caution and you should be liaising with your aggregator around safe usage and safe providers.
If brokers follow safe protocols, do their due diligence and raise any concerns with their aggregator around client or documentary authenticity, you will be better protected against losing accreditations, suffering financial penalties or worse. AI is here to stay and so we need to ensure that we remain vigilant about how it can be used by bad actors. For this, you should always ensure you are dealing with a real person, that you are verifying income and expenses with more than one source of validation, and if something doesn’t feel right, call it out.
Brokers should always be able to evidence where their supporting information has come from and should never accept loan critical supporting information from anyone but the source, the applicant. This measure will go a long way to protecting you and your business should a client attempt to take advantage of you."
Founder and chief executive officer at asset finance and fintech firm LoanOptions.ai
"It is a very dynamic space at the moment, with many of the tools we develop having AI solutions to circumvent, before we then enhance to block the new circumventions they create. We have built a proprietary payslip fraud technology that helps our business stay protected; it flags many indicators of fraud after having seen hundreds of thousands of payslips through our business. We are also able to make this platform available to the broader market, lenders, aggregators and brokers so everyone can be protected.
Brokers need to be careful because we are seeing rampant fraud for payslips and financial documents from generative AI. But also, even more recently, 'selfie ID' images are also being generated as well. We are currently building some solutions to detect AI generated Selfie ID's."
State manager for NSW and ACT home loan broker sales at Macquarie Bank
"Brokers, banks and every other participant in the home lending ecosystem have a shared responsibility and interest in detecting and preventing fraudulent applications. Although banks are constantly on the lookout for fraud, brokers also play an important role in detecting any red flags from the outset. Brokers can help to protect themselves and their businesses by knowing their customers, and most importantly, staying alert. Cross-referencing information to spot inconsistencies in applications, asking more questions, verifying key data points and staying informed about emerging threats can make all the difference.
Nearly all customers and brokers do the right thing and act honestly when applying for a home loan. Unfortunately, there are a small number of fraudulent applications that do occur, and in rare cases this involves broker misconduct. Banks have strong, well-established controls in place to identify and respond to fraudulent activity and these continue to evolve. Although AI document production is an increasing risk, on the other hand, banks are also able to use AI to strengthen our ability to detect, prevent and act on fraud with greater speed and precision.”
Founder and managing director at non-bank lender FinStreet
"Increased fraud is something we’re seeing play out very quickly across the non-bank and private credit space. There are a few key points brokers should be aware of. For one, fraud is becoming more sophisticated, not just more frequent. AI is lowering the barrier to creating highly-convincing documents. The risk today is no longer obvious fraud, but well-constructed applications that appear completely legitimate. In addition, trust is shifting from documents to systems. Brokers can no longer rely on documents alone. The focus needs to move towards consistency across data, transaction logic and whether the overall story holds together. In short, trust is becoming system-based, not document-based. This means shifting from static verification to dynamic validation, one-off checks to continuous monitoring, and human judgment alone to human and system intelligence.
Also, AI should support brokers, not replace them. The real value of AI is as a second layer of validation: flagging inconsistencies, surfacing anomalies and making decisions more accountable. At the same time, it reduces manual work across onboarding and verification, allowing brokers to focus on higher-value areas like client relationships and deal structuring.
Finally, execution discipline matters more than ever. In a faster, more competitive environment, the biggest risk is over-confidence. A strong process is what protects brokers, especially in private credit, where a 'yes' must truly mean 'yes.'"
Chief technology officer at Heartland Bank
"AI is not a substitute for human review and expertise. Accountability still sits with the broker, not the technology. Brokers should verify information. AI can sound confident and authoritative, but its outputs aren’t guaranteed to be accurate. Clients may have AI‑generated research, or advice that looks credible, but is incomplete, outdated or wrong. Brokers still need to verify information against trusted sources and current lender policy and remain responsible for the guidance they provide.
Also, use AI as support, not as an expert. When brokers use AI themselves, it should be treated like a junior assistant rather than a decision‑maker. Anything produced by AI should be reviewed, questioned and validated before being relied on.
And, privacy is critical when using AI tools. Brokers should assume anything entered into an AI system could be stored or reused. And client names, financial details or personal circumstances should never be included. At the same time, AI is making fraud harder to detect, with more convincing fake documents, synthetic identities and scam requests. Brokers should watch for inconsistencies, unusual urgency and situations that don’t quite stack up. And slow processes down if something feels off."
Chief commercial officer at Bluestone Home Loans
"The key message is that AI can be a helpful tool. But it doesn’t replace strong fundamentals. For brokers, protecting themselves comes down to a few practical things: understanding emerging fraud patterns, maintaining robust verification processes and leaning on lender partners who combine technology with experienced human oversight. It also comes down to taking the time to understand their customer, their circumstances and their requirements."
Co-founder and director at Melbourne-based Duo Finance
"Fraud has always been a concern in the banking and finance industry, and the introduction of AI has the potential to challenge that even further. Incorporating AI into your everyday work can offer many benefits, but we must remember that in our broader industry, the integrity of our client’s data is paramount.
To minimise risk, the conversations we've been having with industry peers focuses on conducting thorough research into AI platforms, trialling new technologies before implementation, and developing strong AI and IT frameworks. We all have our preferred AI programs, but it's essential to understand what data can and can't be uploaded to public platforms. There are many finance-specific platforms available that can provide the benefits of AI in a more secure environment."
