There are two pieces of news in the last couple of weeks that relate to brokers and AI. Firstly Lendi Group announced a fully AI-native operating model with agentic tools across its 1,300 plus brokers. Secondly the FBAA told independent brokers not to upload client or company data into public AI tools.
Thinking about these two things together highlights the challenge for smaller and independent operators. This is that the market and customers expect AI-speed service but the easiest and obvious tools are off-limits for anything touching a client file.
So how can these brokers meet this challenge and not have their business swallowed up?
Well the first thing is to recognise the real IP that the brokers have which is the trust their customers have in them and that this is very personal. There are a lot of customers out there who wouldn't dream of changing their broker and we should make sure we don't give them an excuse to.
The power of agentic AI is being demonstrated everywhere. Lendi's AI-native plan has 70% plus of dev work going to AI products that drive its revenue growth and improving margins; LoanOptions.ai, a Sydney AI-native brokerage, cut its full loan application process to under an hour; CBA reports AI-driven credit-decision times for SME loans up to 5 million dollars improved by around 30%, and its business lending book grew 13% to 180 billion dollars.
This presents a massive challenge to traditional brokers.
It's important to note that FBAA is not anti-AI but that it's anti-poor data security and pro-customer privacy.
For smaller and independent brokers though you don't need a big AI strategy or roadmap. A private AI agent pointed at a clean, well-defined workflow beats a public chatbot pointed at a messy pile of files, on both speed and compliance. There are a number of well understood use cases which can be solved quickly:
- Document collection and verification on a new application
- Expense categorisation from bank statements
- First-pass serviceability data assembly
- Post-settlement follow-up
For security these documents sit inside a private controlled environment which aligns with FBAA and legal obligations, and what that setup actually looks like comes down to the workflow layer, where the data lives and how the model is contracted. You also get to determine where the human sits in the process. This is critical for accountability but also for ensuring that you maintain those personal relationships that are so important to your business.
Independent brokers actually have an advantage here, which is their ability to move fast. Quick decision making and powerful new tools means you can leverage agentic AI in weeks rather than years.
If you want a hand working out which workflow to automate first without putting client data at risk, email me at pete@autocognition.com.au.
Not sure which broker workflow to automate first?
Get your free Broker AI Audit. We'll map where your client data flows, which steps are costing you turnaround time, and what can run privately without a compliance headache.