Plaid launches AI credit & fraud models for lenders

Plaid has released new artificial intelligence models for credit, fraud and payments, including an updated version of its LendScore underwriting model.

The rollout includes Instant Link for credit applications, LendScore 2, LendScore Arc, a new fraud model for Protect, and updates to payment risk tools Signal and Guaranteed Payments.

According to Plaid, LendScore 2 delivered 42% stronger predictive performance than traditional credit data alone. The model uses cash flow data to assess a borrower's ability to repay, rather than relying only on conventional credit scores.

Plaid also introduced Instant Link, which allows borrowers who connect their financial accounts to Plaid Consumer Reporting Agency to consent to sharing cash flow insights for future credit applications. Lenders can access those insights in under two seconds, the company said.

Credit models

The expanded LendScore suite includes specialist versions for different parts of the lending market. Plaid said LendScore 2 Auto reduced delinquency by 26% among deep-subprime applicants at the same approval rate, while LendScore 2 Home Lending approved 6.3% more borrowers at the same level of risk.

LendScore Arc, a separate model, is Plaid's first transformer-based credit risk score. In early testing, it learned from the order and timing of a borrower's transactions and produced a 20% predictive lift on deep-subprime borrowers and a 24% lift on superprime borrowers compared with Plaid's core model, according to the company.

Plaid positioned the new products as a response to the limits of traditional credit scoring for many consumers. The company said millions of adults in the US fall outside the reach of conventional credit assessment and that cash flow data can give lenders a broader view of income, spending and bill payments.

Will Robinson, Chief Technology Officer at Plaid, said the models draw on data patterns across the company's financial network.

"Our credit, fraud, and payments models are unique because they bring deep financial context to every problem they're solving. They build on foundation models that already understand how financial behavior unfolds over time, across the Plaid Network, and that means better decisions, and better outcomes, for our customers and the millions of people who depend on those services to manage their own financial lives," Robinson said.

Michelle Young, Credit Product Lead at Plaid, said the changes are intended to make cash flow underwriting faster and easier to use in lending decisions.

"Cash flow data tells a more complete story about a borrower. The challenge has been turning that story into insights that lenders can access and act on with speed and confidence. The next generation of LendScore and specialized models close that gap at scale, and with Arc, we're giving lenders new tools to expand access to more affordable credit," Young said.

Fraud and payments

Beyond credit, Plaid has added a new AI foundation model to its fraud product Protect. According to the company, the model is trained on patterns and hundreds of millions of data points from its network and examines sequences of events rather than a single snapshot.

Plaid said the fraud model strengthens Protect's Trust Index scoring framework. In internal evaluations, it delivered up to 40% relative improvement over previous baselines.

The company also said its sequential foundation model now supports Signal, its ACH payment risk model. Plaid said the system reads an account's transaction history in sequence to assess payment risk more accurately.

In testing, the updated model helped Signal prevent 26% more ACH returns without increasing false flags, according to Plaid. Guaranteed Payments uses the same risk analysis to offer businesses approval options including delayed release and partial guarantees, rather than a binary approve-or-decline outcome.

The San Francisco-based group said its network connects to more than 12,000 financial institutions across the US, Canada, the UK and Europe. Plaid added that more than 9,000 companies use its products across payments, lending and fraud prevention.