AI is moving at a pace that can leave regulation, infrastructure and even internal teams playing catch-up. But that doesn’t mean that lenders should slow down. Lenders need to be smarter about where AI should be introduced.

We had a chat with Mario Husha, CTO, Lovey at FTT AI Transformation in June.  Mario shared how AI is already finding a place across the SME lending journey, from origination and underwriting, through to payout. But for Mario, innovation is not about handing the entire process over to machines.

While AI is increasingly ready to support many stages of the journey, particularly where human oversight remains in place, there is still a clear boundary around decisions involving money movement. More importantly, borrowing is rarely a purely transactional decision. Behind every SME loan is a business, a person, and a reason for needing finance. That human element still matters.

For financial institutions navigating the build-versus-buy question, Mario argues that the answer starts much closer to home. Rather than immediately looking outward for the latest AI solution, organisations need to build confidence internally first. This means creating a culture where engineers and the wider business understand what the technology can do, why it is being introduced and where it can genuinely add value. Once that foundation is in place, the decision to build internally, buy an existing solution, or partner with another organisation becomes much clearer. The technology may be changing at speed, but the principle remains straightforward: successful AI adoption starts with people, not products.

That same principle extends to the next generation of collaboration between financial institutions and technology providers. Mario sees AI agents, MCP and evolving approaches to agent communication as the next step in a familiar story: getting different systems to communicate effectively. APIs and data sharing have already transformed how financial services connect, but the focus is now shifting towards what capabilities can be made available to AI agents and how easily those agents can interact with them. For SME lending, the opportunity is significant, but the foundations still matter. Clear systems, accessible capabilities and the right information in the right place will be critical to making increasingly autonomous AI useful rather than simply impressive.

Watch the full interview to hear Mario’s take on where lenders should be moving fast, where they should pause, and what needs to be in place before AI can truly transform the lending journey.

And for more on the fast-growing impact of AI tools and technologies across the financial services sector and beyond, join us at the next FTT AI Transformation, which runs on 9-10 November in London.