Plaid Launches AI Models for Credit, Fraud and Payment Risk
Plaid introduced AI models for lending, fraud detection and payment risk, according to Finextra. The models analyze cash flow and transaction sequences to support decisions by financial firms.
LendScore 2 uses cash flow data to assess a borrower's ability to repay. Plaid says it was 42% stronger at predicting repayment than traditional credit data alone. Its transformer based LendScore Arc model produced a 20% lift for deep subprime borrowers and a 24% lift for superprime borrowers in early testing.
Plaid also rolled out a foundation model for fraud detection, trained on hundreds of millions of data points from its network. Internal evaluations showed up to a 40% relative improvement over previous baselines.
The same sequential model powers Signal, Plaid's ACH payment risk system. In testing, it prevented 26% more ACH returns without increasing false flags.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like Finance AI Weekly or Daily AI Brief.
Also, consider following us on social media:
More from Finance
Nov 18 Payhawk raises $280 million at a $2 billion valuation to build an AI-agent finance platform Nov 18 Visa expands Agentic Ready program to more than 50 payment partners for AI agent purchases Nov 18 Binance Wallet opens campaign for tokenized shares tied to Anthropic's IPO Nov 17 CommBank releases free CodeCheck tool that uses Codex and Claude Code to find software vulnerabilities Nov 17 BNY and Microsoft complete first AI agent negotiation of a US Treasury tradeFinance AI Weekly
Weekly newsletter about AI in finance. Covers AI-driven trading, fintech innovations, and data analytics transforming markets
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
UK credit risk data company Planky raises €6.9 million to spot borrower stress before defaults
Pliant raises $80 million for AI agent payments
Plinto Raises €2.9 Million for AI Financial Controls
Spreedly adds Riskified fraud checks for orders placed by AI shopping agents
Chargeflow launches Prevent 2.0 for Shopify fraud screening
Daily AI Brief: the AI news that matters, in your inbox.