Directorio Digital: Web Directory and Local Businesses

Credit agencies and lenders weigh AI as a new way to judge who can borrow

Lenders and credit firms are testing AI to assess borrowers beyond traditional credit scores. Supporters cite wider access to credit, while critics warn of hidden bias and decisions that are hard to challenge.

Directorio Digital: Web Directory and Local Businesses · · 2 min read · Computers and Internet

Artificial intelligence is moving into the business of deciding who can borrow money, according to a report by BBC News. The report looks at whether automated systems could do a better job than traditional credit scores when lenders weigh up applications for mortgages, car loans or phone contracts.

Limits of the current system

In the UK, credit scores come from three firms, Experian, Equifax and TransUnion, which draw on broadly similar data. Critics quoted in the report say these scores mainly reflect past behaviour and may say little about what a person can afford today. People with little borrowing history, such as young adults and recent arrivals in the country, can be left out altogether. Consumer finance lawyer Funmi Olufunwa told the BBC that the country is overly focused on these scores.

AI as an alternative

In the United States, firms such as Upstart and Zest AI already sell AI-based assessment tools to lenders. Upstart says its machine-learning models examine more than 2,500 variables and that this lets it approve more loans across demographic groups. The established UK agencies are also experimenting. Experian announced a partnership with OpenAI in August 2026 so users can check their score inside ChatGPT, Equifax has launched a machine-learning platform for credit and fraud checks, and TransUnion said in March it was integrating Google's Gemini models into its data products.

Not every newcomer is embracing the technology. Noggin HQ, a credit reference agency authorised by the Financial Conduct Authority in June, analyses current account transactions to help lenders but deliberately avoids AI, citing the risk of built-in bias.

Concerns about bias and transparency

Sandra Wachter, professor of technology and regulation at the University of Oxford, warned that AI trained on past lending decisions may simply reproduce existing inequalities. She argued that firms must apply current rules and add new safeguards, saying that without them such systems would "fortify, foster, and enhance discrimination." Others stressed that consumers need to know what data is used about them and be able to dispute errors. Olufunwa also suggested younger people may accept this use of their data more readily than older generations.

What it means for technology users

For people who follow computing and online services, the story shows how AI tools are spreading into everyday financial decisions, often through partnerships between data companies and large technology platforms. It is a reminder to understand which personal data apps and services can access, to check credit records regularly, and to look for clear explanations and appeal routes when an automated system makes a decision.

Source: BBC News

This article was written with the help of AI, based on reporting by BBC News.

More Computers and Internet news