Practicality over hype: AI comes of age at AFC UK Summer Conference

A real-world reset on AI in lending

The AFC UK Summer Conference 2025 brought together an engaged audience of finance professionals eager to cut through the noise around artificial intelligence and uncover practical steps forward. In the tech wrap-up session, sponsored by Lendscape, panellists distilled a day’s worth of FinTech and AI-focused sessions and workshops into a clear narrative: AI is ready for business – but only if applied with purpose and pragmatism.

Led by Steve Taplin, Managing Director at Lendscape, the session featured candid and constructive contributions from Richard Huston (VAMOS), Cat Powell (Novuna Business Finance), and Andy Trimmer (Simply Asset Finance). Each brought a dual-lens view – blending technical insight with strategic awareness – on how lenders can responsibly and effectively bring AI into production.

Key takeaways: from hype to hands-on

1. Start with the business problem, not the technology

“AI is a tool, it’s not a silver bullet,” said Andy Trimmer, reminding attendees that throwing AI at every problem is a costly and inefficient strategy. Instead, the consensus was to focus on clear, measurable problems: repetitive manual tasks, inefficient document processing, slow customer response times.

“Identify areas of inefficiency,” echoed Cat Powell, “and ask yourself: where can AI give me speed or structure that delivers value today?”

2. Proof of concept without ROI is just a science project

A recurring theme was that AI pilots often stall because they aren’t designed with a clear outcome in mind. The day’s poll confirmed this: 44% of attendees cited lack of clear ROI measurement as the main barrier to scaling AI pilots.

“If your POC doesn’t include success criteria and a way to measure impact,” Powell warned, “you’re already setting it up to fail.”

3. Document processing and customer service: the low-hanging fruit

The two most popular AI priorities – document processing and customer service automation -reflect a shift from abstract ambition to concrete application.

“These are not future-state ideas,” said Trimmer. “If 33% of you voted for document processing, good news, it’s ready today.”

Meanwhile, Huston noted that customer service AI is making a difference not by replacing people, but by assisting them: “Internal support tools are where many are starting – freeing staff up from repetitive questions so they can focus on higher-value work.”

4. Buy vs build? It depends but stay involved either way

Whether outsourcing or developing in-house, maintaining internal involvement is critical.

Powell noted that some of Novuna’s slower-moving projects were ones where external teams lacked domain knowledge. “The context of your business matters,” she said, “and if you’re going to build, make sure you’re also investing in people to maintain and evolve those solutions.”

Trimmer agreed, highlighting that AI is now embedded in many existing tools (like Zoho or Microsoft), and organisations should start by exploring what’s already available within their tech stack.

5. The skills gap is real, but not a dealbreaker

Only 17% of attendees felt “well-equipped” internally to deliver AI, while a third admitted they would need significant external support. But that’s not a weakness, it’s a recognition of how fast the landscape is evolving.

“You don’t need to have AI experts overnight,” said Huston. “But you do need curiosity, openness to experimentation, and people in your team who are willing to learn.”

Upskilling, internal collaboration, and a flexible tech strategy were emphasized as more important than any single hire or vendor.

Measuring AI’s impact: not just hours saved, but quality improved

Despite the temptation to chase hard numbers, several panellists reminded the audience that AI’s value often includes qualitative improvements: better customer experience, faster employee workflows, and higher satisfaction.

“If you’re asking people to trust the tech with data or customer journeys, you need to show the business value in clear terms,” she said.

Caution, but not complacency

As the session drew to a close, Taplin asked the panel to reflect on the broader implications of widespread AI adoption, including environmental impact and job displacement. While concerns were acknowledged, the group largely agreed that responsible implementation can amplify human talent rather than replaceit, and that many fears are inflated by misinformation.

A shift toward confidence and action

In the final poll, 56% of attendees said they now feel more prepared to implement AI, with the rest holding steady. That’s a significant shift compared to just a few months ago.

And perhaps that’s the real takeaway: AI is no longer just something to watch – it’s something to do. The key is to do it strategically, measurably, and responsibly.