iscover 
In the latest episode of Broker Daily Uncut, Alex Whitlock was joined by Finni brokers Eva Loisance and Costa Arvanitopoulos to unpack how AI is being integrated into daily workflows and whether it poses a long-term threat or benefit to the broker model.
The conversation was sparked by the news that fintech Quickli has launched two new AI tools. Jiffi AI, Quickli’s in-built AI policy and product research tool, provides instant answers from 45 lenders, while Doc Renamer automatically detects and renames client documents at scale.
“We’ve been hearing about AI for broking for quite some time and I’ve always been a bit disappointed when I’ve looked at it,” Loisance said.
“But last week, Quickli actually came up with quite a few really good things. We’re able to ask Jiffi difficult scenarios and it gives us quite a bit of insight.
“You can look at all the policies and go the old fashioned way, pick up the phone and hope the business development manager I need is picking up the phone to give me the right information. That would take hours.”
Arvanitopoulos added: “It’s our job as brokers to find solutions for clients. But things like this, it just makes life so much easier. It cuts out a 10 hour job to one hour, and then that other nine hours you can spend nurturing other clients.”
AI can also help bring together large amounts of information that would be unrealistic for a broker to remember entirely.
“I’m not someone that remembers policies forever,” Loisance said.
“Three weeks later I’ll probably have to ask the same question. Tech can take that pressure away.”
Arvanitopoulos said: “Because you’re dealing with different scenarios and different structures, you need to find a lender that has a policy that fits that specific person. So much can change from the assessment process that you can’t remember all policies. It’s just not possible. And everyone’s niches and stuff like that. AI just makes this easier.”
Beyond policy matching, AI is also being used to assist with servicing calculators, SMSF lending calculations, and low-doc scenarios – areas traditionally requiring manual comparison across multiple lenders.
Arvanitopoulos and Loisance said that they found some lenders further ahead than others. Macquarie Bank was flagged as a bank with fast and streamlined turnaround times.
Will brokers get replaced?
While AI is being embraced as a productivity tool, the question of whether brokers themselves could be replaced remains open.
Simple, “vanilla” deals, such as straightforward first home buyer applications, were identified as the most susceptible to automation, particularly where structures are uncomplicated, and policy alignment is clear.
However, more complex transactions involving investors, SMSFs, trusts, or layered structures were described as far less likely to be fully automated in the near term.
“Good tech impacts everybody,” Whitlock said.
“Some lenders are delivering it in the right way which is clearly transformational to brokers. But I don’t think that tech will ever replace brokers.”
The role of judgement, speed of turnaround, lender relationships, and negotiation, particularly in time-sensitive settlements, was highlighted as an area where human brokers continue to play a central role.
“This is the beauty of the complexity of being a broker because it isn’t just, it’s not a matrix or cheapest loan,” Whitlock said.
“There’s a whole host of things which automation can never do.”
Best interests and the human touch
The brokers flagged two other issues with AI moving forward.
The first concern was that the tightening of expectations around best interests duty (BID) is increasing documentation requirements and scrutiny around product selection.
Brokers described a growing need to provide clearer written rationale for lender and product choice, particularly where a recommendation is not the lowest rate available.
While AI is already being used to assist in drafting BID commentary, Whitlock raised questions about accountability in a fully automated environment.
He said: “Automation can’t govern automation. If you end up with complete robo-broking there’ll be no oversight.
“And then, who’s going to own all the automation? You imagine it would go back to the banks or to big faceless organisations. There is an accountability with brokers that cannot be automated.”
The other irreplaceable aspect of brokers that was flagged was the human touch.
“I think there’s just something about borrowing a lot of money that is so fundamentally human,” Whitlock said.
“You’ve got customers who want warm blooded human beings that want to deal with a broker. On top of that, you’ve got scrutinisation of that human being that’s accountable to the customer.”
[Related: Cynario and Salestrekker partner on AI innovation]