Oakbrook Loans builds UK lending's first Generative AI credit scorecard

Oakbrook uses AI to make more accurate, fairer lending decisions by identifying patterns in credit data that traditional methods miss.

Every year, thousands of people are declined for a loan not because they aren't creditworthy, but because the model assessing them doesn't see the full picture.

Oakbrook has built an AI powered scorecard that changes that.

Developed by Oakbrook's Data Science team, the scorecard is the first in the company's history built using Generative AI, and one of the first of its kind in UK consumer lending. It is designed for everyday borrowers seeking personal loans up to £5,000: people who fall outside the mainstream credit market not because they are reckless with money, but because life rarely goes in a straight line.

Built for the people credit scoring gets wrong most often

Near-prime borrowers are among the hardest to assess accurately. A redundancy, a medical bill, a difficult year, any of these can leave a mark on a credit file that does not reflect who someone is as a borrower today. Traditional scoring models struggle with this. They work from predefined features, and they can only see what they have been told to look for.

This scorecard works differently. It uses Generative AI to read raw credit file and application data and surface patterns that conventional feature engineering would either miss entirely or take significantly longer to find. The result is a richer, more complete picture of each applicant, one that gets closer to the reality of their financial life rather than a simplified version of it at a point in time.

Jamie Ownsworth, Head of Data Science at Oakbrook, said:

"Behind every application is a person trying to improve their financial situation - cover an unexpected cost, get some breathing room, consolidate a debt. They deserve a decision that reflects their situation, not one that fails them because the scorecard doesn’t see past the surface.

Our new scorecard narrows that gap. What made it possible was using AI to analyse the same raw data and finding signals in it that traditional methods had always missed. I'm proud of what the team has built, and I'm confident it will mean better outcomes for the people we're here to serve."

What changes for borrowers

A more accurate scorecard means more personalised decisions. Creditworthy borrowers who might previously have sat just below a cut-off have a better chance of being seen clearly. People who would genuinely struggle to repay are identified sooner, before a loan makes their situation harder.