Lending Briefing: The ethical use of alt data and AI could pave the way to a fairer financial system
- While AI and alternative data have been around for a while, the industry has only begun to scratch the surface regarding their potential to unlock a fairer financial system.
- If these tools are used ethically, the results can be beneficial for both the lender and the borrower at scale.

While AI and alternative data have been around for a while, the industry has only begun to scratch the surface regarding their potential to unlock a fairer financial system. If these tools are used ethically, the results can be beneficial for both the lender and the borrower at scale.
Having better data, both qualitative and quantitative, and pairing that with the appropriate underwriting technology can unlock new opportunities for lenders, and give access to credit for segments that were previously underserved.
Lenders can safely stay within their risk appetites while allowing technology to help them widen their customer base. And it looks like they've started to take note of that.
"There seems to be a much greater appetite and openness from lenders to look at alternative data as a way to further enhance their risk assessment tools", according to Ron Benegbi, CEO and founder of Uplinq, a global credit assessment platform focusing on SMB lending.
The goal is to present a more accurate picture of the true credit risk for that potential client. Companies like Uplinq are targeting the small business sector, a historically underserved corner of the market, using a wide range of data points – from environmental factors and local market economy all the way down to street level traffic patterns and cell phone usage. All this data gets aggregated with traditional financial information and bureau scores to add the final pieces of the puzzle.
Source: Argyle
If used ethically, new data streams and software can help pave the way towards a fairer financial system. They can show that financial inclusivity doesn't have to be a compromise. A lender is not taking more risk by underwriting a new segment - in fact, it's losing a potential revenue stream by denying credit to that segment in the first place.
"We're just a small part of that solution in terms of taking the bias out of the formula or equation to help lenders make decisions based on data and science and not based on ethnicity or anything else for that matter," said Uplinq's Benegbi.
And ultimately, this can also have a positive externality in terms of fighting discriminatory lending practices. BetaBank is a great example of this – a fully digital Black-owned bank built to serve SMBs that would traditionally get refused and fill the gap left behind by big banks.
BetaBank has paid close attention to how they structured their underwriting models, to weed out the bias. This is where the issue of ethics comes in, because if your assumptions are garbage, then the output of your model is going to be garbage, CEO Seke Ballard told Tearsheet.
"When training any algorithm, you are using data that is generated through human activity, and that data often reflects the bias of that human activity. So if you are not intentional about rooting out that bias in human activity, then the output of your model is simply going to be the same thing that humans do, except it’s going to be a black box, and no one can really tell what’s happening in that black box," said Ballard.
Zest AI is another good example, as an AI-powered underwriting software that aims to provide more accurate and inclusive lending insights.
Zest AI recently raised $50 million in an investment round which saw participation from a number of credit unions – Suncoast Credit Union, Golden1 Credit Union and Hawaii USA Federal Credit Union – all working with Zest AI to modernize their lending operations.
"I think the time for change is long overdue to transition away from this old math that served us well in the 1950s when they started, but there is a new way. And that new way is to leverage machine learning," said CEO Mike de Vere on a Tearsheet podcast.
What we’re reading
Small business lender Capital On Tap raised a $110 million revolving credit facility
AI-based auto finance platform Lendbuzz secures a $150 million credit facility from JP Morgan
Credit report and compliance solution provider Credit Bureau Connection announced its acquisition of credit soft pull provider CreditDriver
Square originated 126,000 loans for a total of $1.14B in Q3
Upstart funded $9M in Small Business Loans in Q3
Klarna's open banking arm Kosma helps SME lender Krea boost loans
Adyen adds small-business credit to take share from banks, fintechs
Chart of the week
Source: McKinsey
What we’re writing
Data Snack: PayPal continues to report slow user growth, focusing on monetization
PayPal beat market expectations on both top and bottom lines on its Q3 results, but much of the focus went towards its stagnant user growth figures, which left the market pondering over the digital payments company’s strategy through the market downturn.
How fintech apps help consumers weather economic challenges, in 4 charts
Fintech adoption skyrocketed in 2020 due to Covid-19 restrictions. Forced to stay home, consumers flocked to finance apps to take care of their banking needs. But, as the lockdown restrictions loosen, are consumers still sticking with fintechs?
Does B2B BNPL have the potential to emerge as the next top fintech trend?
The proliferation of the B2C model would suggest that it is only natural for businesses to put on their thinking cap and expand this payment option to business buyers – but will it taste the same level of demand and success?