Business

Should AI judge our creditworthiness?

Hope’s experience shows how important a credit score can be, but, like her, it doesn’t tell the full story.

“As a nation, we are obsessed with debt,” said Funmi Olufunwa, a consumer finance attorney and founder of financial education business, Hoops Finance.

Three firms – Experian, Equifax and TransUnion – produce credit scores in the UK, and they all work with almost the same data.

“Debt looks back. We look at what happened in the past, which does not mean what will happen in the future,” said Ofunwa.

“But if someone wants to lend you money, they look at what is really affordable,” he said.

Traditional credit scoring systems also disadvantage “small file applicants” – young people and new immigrants to the UK who may not have built up a credit history.

But with more financial data available than ever before – and a growing desire for AI to make sense of it – can artificial intelligence help lenders identify who is safe to lend to?

In the US, loan decisions will often be based on a credit score from a company called FICO, which uses data from the three major credit reporting companies.

But AI-powered lending firms, Upstart and Zest AI have both built businesses that give lenders more options when it comes to checking a customer’s creditworthiness.

“Thanks to our machine learning models that look at more than 2,500 variables, we believe we can lend more money at lower rates to borrowers of all demographics,” Upstart said in a statement.

Mortgage expert and founder and CEO of Mortgage Mum, Sarah Tucker says: “AI has the potential to help lenders analyze vast data sets, quickly identify patterns and build an individualized picture of affordability and risk.”

If technology allows lenders to better understand so-called small-file applicants, rather than simply applying strict rules, Tucker believes there is a “real opportunity to improve financial inclusion”.

But like many other applications of AI in everyday life, there are concerns.

“Consumers need to understand what information is being used to make decisions about them, and they need to be able to challenge something if it’s not right,” explained Tucker.

Credit scores have been a controversial topic for decades for this reason, says Sandra Wachter, professor of Technology and Law at the University of Oxford.

“We don’t check whether past decisions were right,” said Wachter, adding, AI is not aware of existing social inequalities that influence decisions and can simply “repeat what it has seen in the past”.

While companies like Upstart and Zest AI see technology as a tool for capitalization, Wachter believes that real change requires accountability, and automation.

He insists that firms must enforce existing laws and establish new avenues of oversight to ensure that AI does not simply perpetuate the biases of the past.

“Not including any safeguards will reinforce, encourage and promote discrimination.”

Upstart says it is fixing the issue.

“We’ve invested in creating objective, automated tests to detect bias, and we’ve incorporated those tests into our model training program so that it can become more accurate and fairer over time,” the company said in a statement.


Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button