Affirm Shifted Underwriting Model to Boost Approvals

The company's new system may make it easier for shoppers with limited credit histories to access point-of-sale financing.

Updated on Oct. 5, 2026 in Economic Indicators

Isometric editorial illustration of geometric blocks representing complex, interconnected financial data layers in oxblood and slate blue. — Affirm reported a 3.4% increase in completed purchases during its fiscal fourth quarter, driven by a new real-time underwriting model. AI Illustration. Upload story photo >

Live Poll

Do you trust artificial intelligence to make fair lending decisions for your personal purchases?

Affirm recently deployed a new transformer-based underwriting model at checkout that assesses shoppers individually in real time. This change helped the company increase completed purchases by 3.4% during the fiscal fourth quarter.

Why it matters

By leveraging existing credit bureau data like payment history and balances more effectively, the system allows the lender to approve consumers who lack traditional FICO scores. This shift aims to improve decision-making accuracy and broaden credit access for users.

Affirm processed 53 million transactions in fiscal Q4 2026, marking 41% year-over-year growth. While total active consumers reached 27.8 million, the average order value declined by 4%.

The players

Affirm

A financial technology company that provides point-of-sale consumer financing and loan services.

The details

The new model identifies patterns within and across consumer accounts to perform real-time underwriting for every purchase. By analyzing established credit behavior, the company can extend financing to individuals who have historically struggled to qualify for credit. This method contrasts with traditional static scoring, as it evaluates unique patterns of activity to facilitate consumer transactions.

Fiscal Q4 2026 saw Affirm process 53 million transactions.
Transactions per active consumer reached 7.0 over the trailing 12 months.

Money Landscape

The adoption of transformer-based modeling by Affirm follows a broader financial trend of using machine learning to expand credit access beyond traditional FICO-based scoring. This shift marks an evolution in how lenders weigh real-time data against historical credit performance.

If you have a limited credit history or no FICO score, these model changes may increase your chances of being approved for point-of-sale financing at retail checkouts. Review your budget to ensure that any new credit lines fit your long-term debt-repayment capacity.

The takeaway

The move toward automated, data-intensive underwriting could grant more consumers access to buy-now-pay-later options. Before utilizing these financing tools, consider whether the purchase aligns with your monthly budget and consult a financial professional regarding your overall debt management.