By
A whopping 80% of broker loan deals have the potential to be handled by artificial intelligence within just five to seven years, according to the founder and CEO of Australian AI online loan matching platform LoanOptions.ai.
Julian Fayad (pictured above), who has been developing AI use cases in broking since launching in 2020, said there is no doubt AI will be deployed to do the broking aspects of lending “in some capacity” in the future.
In the near term, AI will enhance brokers’ capabilities and efficiency, he said; it will decrease the amount of friction for customers and allow them to access more loan products more directly.
However in the longer term, which Fayad defined as a five-to-seven-year time horizon, he said the technology could have the capacity to handle 80% of the broker market’s current lending deals.
“If you've got a vanilla transaction that is inside the box, AI can do that faster, more reliably, without any bias – or limited bias, I should say – comparatively to a human,” Fayad said.
Fayad said a standard refinance, or a simple personal loan with no car dealership involved, were examples of mildly complex decision-making that AI can do “at a scale and speed humans can’t”.
“It also doesn't have sick days and close on public holidays and all that kind of stuff,” he said.
Whether AI can be deployed to handle this level of transaction volume will depend on regulation. With few regulatory obstacles, Fayad said it could happen as soon as three years from now.
With more regulatory obstacles, it is more likely to take between five and seven years, as regulators get comfortable with AI in relation to consumer protections like responsible lending obligations.
The federal government, for example, is still finalising new mandatory guardrails for AI in high-risk settings; it is yet to fully define high risk, where it will require humans be kept “in the loop”.
“Brokers will need to supervise AI deals until regulators, including ASIC, are comfortable AI is as good or better than a human and that it is not going to put customers in harm’s way,” Fayad said.
Brokers are likely to be heavily involved in building and deploying AI technology, and will be driven by the desire to “assist more clients get better options and outcomes at a faster pace”, Fayad said.
“There will be lots more efficiency, transparency and speed throughout the process.”
Fayad said brokers should consider whether the type of business they were chasing, or the way they positioned their business, could be impacted by being easily machine automated.
“If you’re positioning yourself for vanilla refinancing deals, where you are just all ‘rate rate rate’, and you are not adding value to customers, computers can do it faster and better than you can.”
Rather than system-generated, boilerplate-style transactions with little human interaction, more complex transactions with multiple parties would take longer to automate.
These could include more difficult first-home buyer deals, investors dealing with complex trust structures, SMSF lending deals or more complex equipment finance transactions.
“That would make sure you have the longest time horizon possible before AI can take over,” Fayad said.
In the long run, he said it made sense for brokers to embrace the benefits of AI technology for customers, in a similar way the market had adopted digital technologies over “paper and pen”.
“The longer you hold on, the more risk you take of becoming extinct,” Fayad said. “At a minimum, [with AI] you are hedging your bets, but there’s a good chance you’re betting on a winning horse.”
