In the following Q&A, IFA Magazine sits down with Zahid Bilgrami, CEO of Mortgage Brain, to discuss the launch of the company’s new AI Charter, the hidden risks of “shiny veneer” AI tools, and why the industry needs more scrutiny around secure AI usage.
IFA Magazine: Zahid, congratulations on the launch of Mortgage Brain’s AI Charter. The reaction from the industry has been very positive. Why do you think it’s struck such a chord?
Thank you, and yes, the response has been genuinely encouraging. I should say upfront that this is a topic I’m deeply passionate about, so much so that I spend a fair bit of my spare time tinkering with my own AI projects at home.
I think it’s struck a chord because, frankly, it’s long overdue. Over the past couple of years, we’ve seen many shiny new AI and tech tools appear in the mortgage industry, all promising faster, quicker, smoother, more accurate journeys and workflows. Brokers, firms and lenders have understandably been caught up in the excitement of what these tools can deliver, particularly against the backdrop of constant pressure on the industry to be faster, faster, faster at what it does.
But excitement isn’t a strategy. The AI Charter gives the industry a framework to slow down for a moment and ask some hard questions about what’s actually sitting under the bonnet of these tools. That conversation needed to happen, and I think people have been waiting for someone to start it.
IFA Magazine: For readers who haven’t yet seen the AI Charter, can you briefly talk us through its four pillars?
The AI Charter is built around four principles that we believe should underpin any responsible use of AI in the mortgage industry: Cost, Intellectual Property, Consistency, and Speed.
Cost is about sustainability and predictability – making sure the AI you’re relying on today isn’t going to become prohibitively expensive tomorrow because a third-party provider has hiked its prices. Intellectual Property is about data sovereignty – where your client data lives, who has access to it, and whether it’s being used to train someone else’s model.
Consistency is about making sure the same question gets the same answer every time, which is non-negotiable in a regulated advice environment. And Speed is about being fit for purpose rather than fast for show – recognising that not every process needs the most powerful AI engine behind it, and that the right tool for the job often isn’t the flashiest one.
Together, those four pillars are designed to help the industry cut through the marketing noise and evaluate AI tools on the things that actually matter.
IFA Magazine: You’ve been quite outspoken about the risks of some of these new tools. What specifically concerns you?
My concern is that a lot of what’s being marketed as cutting-edge proprietary AI in our sector is, in reality, little more than a shiny veneer wrapped around a third-party large language model (LLM) – ChatGPT or something similar. That’s a very different proposition from what’s being sold, and it carries significant risks that I don’t think brokers and firms are being made fully aware of.
We operate in a regulated industry. Brokers and lenders need to be absolutely clear about how the data they’re sharing with these tools is being used, where it’s being processed, what guardrails sit around it, and what happens if something goes wrong. When you’re routing sensitive client data through a public LLM, those questions often don’t have comfortable answers.
I could see the problems coming down the line if we didn’t tackle this head on. The AI Charter is our attempt to do exactly that.
IFA Magazine: Let’s talk about the audit trail issue. Why is this such a critical point for regulated advice?
This is the point I’d urge every broker and compliance officer reading this to really sit with. If a broker is making consumer-facing decisions and giving advice based on the outcomes of an AI tool, they need to be certain they can follow the audit trail as to how they arrived at that decision. That’s a baseline regulatory expectation.
The problem is that public AI LLMs are probabilistic by design. Ask the same question twice and you’ll often get two different answers. That means the audit trail becomes fuzzy – it stops being a clean line you can walk a regulator or a client down. For a regulated advice journey, that’s simply not fit for purpose.
There’s also a jurisdictional dimension that often gets overlooked. Data entered into a public LLM can be processed in the US or elsewhere, entirely outside of UK or EU jurisdiction. That creates a whole issue around compliance and regulation that many firms haven’t fully grappled with and it’s not something you can wave away with a tick-box on a procurement form.
At Mortgage Brain we’ve built AI that returns consistent, deterministic outcomes. The same input produces the same output every time – a fundamental requirement of operating in a regulated environment.
IFA Magazine: Cost is the first pillar of the Charter. Why did you decide to put it front and centre?
Because it’s the issue that’s going to bite the industry hardest, and soonest, and almost nobody is talking about it honestly.
The reality is that a lot of the AI tools being sold into our sector right now are running on very large, general purpose models built by OpenAI, Google or Anthropic. They’re impressive pieces of technology, but they’re also expensive to run, and the pricing is set by the provider, not by the firm using them, and certainly not by the broker at the end of the chain. As the AI market matures, the consensus view among analysts is that those prices are going to rise. When they do, those costs get passed down the line.
So, a firm that today is paying a comfortable monthly fee for an AI-enabled tool could, within a year or two, be facing significantly higher costs, not because the tool is any better, but because the provider behind the provider has changed its pricing. That’s a real commercial risk, and firms should be factoring it into their decision making now.
IFA Magazine: And how does Mortgage Brain’s approach address that?
Because we build and run our own AI, we control the infrastructure. We’re not exposed to somebody else’s pricing decisions, and we don’t have to pass sudden cost increases on to our customers. We can commit to predictable, sustainable pricing over the long term, which in a regulated industry where firms need to plan their cost base carefully, is genuinely important.
There’s also a point about efficiency here. Using one of the very large general purpose models for every mortgage task is, to use the analogy from the AI Charter, a bit like using a jet engine to power a bicycle. It’s overkill, and you’re paying for that overkill. We build smaller, focused AI systems designed for specific jobs, which means lower running costs and more predictable pricing, without any compromise on the quality of the outcome.
IFA Magazine: The Charter sets out a list of questions brokers should be asking their tech providers. What are the red flags and green flags firms should be looking for?
The starting point is not all tech and AI tools are created equally, and brokers should be asking the questions to prove that for themselves rather than taking marketing collateral at face value.
The questions we’d encourage people to put to their providers are fairly straightforward covering such issues as data storage and processing, costs per transaction, hacking threats, and data ownership.
A provider that can answer such questions clearly, in writing, is showing green flags. A provider that gets vague and can’t come up with the specifics, is showing red flags. That’s exactly why we’ve developed a list of questions for brokers, available in our new AI Zone, so that firms can enter into agreements with their eyes wide open or, at the very least, hold their existing providers to account.
IFA Magazine: Is this just a Mortgage Brain passion, or do you see this as something bigger?
This is much bigger than just Mortgage Brain, and I want to be clear about that. This is an industry issue, and we’re proud to be leading the way on it, but the AI Charter isn’t a marketing document dressed up as a principles piece. It’s a genuine attempt to raise the standard for everyone.
We’ve been in this industry a long time. That experience and that depth of knowledge means we can see what’s coming down the line – the regulatory scrutiny, the data protection challenges, the consumer outcomes questions – and we have a responsibility to speak up about it. If we don’t, the industry risks sleepwalking into problems that were entirely avoidable.
IFA Magazine: How does Mortgage Brain’s own approach differ in practice?
The fundamental difference is that we build our own AI and tech. Broker data goes no further than us. It isn’t being handed off to a third-party LLM provider in another jurisdiction. That alone resolves a huge category of risk.
On top of that, we’ve developed AI specifically designed to return consistent outcomes, rather than the probabilistic outcomes you get from public LLMs, which are simply no good for regulated advice journeys. And because we control our full AI stack, we can also commit to predictable pricing over the long term. We’re not exposed to sudden cost increases from an external provider that then get passed on to our customers.
In short, we’ve built AI as infrastructure, not as a veneer. That’s the distinction the AI Charter is trying to draw out for the wider market.
IFA Magazine: And finally, what would you say to a broker or firm reading this who’s now starting to wonder about the tools they already have in place?
I’d say ask the questions. Don’t wait for a regulator, or a data breach, or a client complaint to force the issue. Sit down with your providers and work through the checklist we’ve put into the AI Zone. If the answers you get are reassuring, that’s fantastic. If they’re not, you now know where your exposure sits and you can do something about it.
This isn’t about slowing innovation down. It’s about making sure the innovation we adopt in this industry is genuinely fit for the job it’s being asked to do. That’s what our clients deserve, and ultimately it’s what their customers deserve too.
Zahid Bilgrami is the CEO of Mortgage Brain
