Which banks lead in AI for fraud detection and credit decisioning?

Which banks run AI for fraud, AML and credit decisioning in production?

The 30 highest of 131 banks with a fraud, AML/KYC, credit-underwriting or fair-lending AI use case on their page, each with its stage. Figure: the first disclosed number on the bank's page about fraud, AML, approvals, losses or alerts (a result where the bank gives one, a loss or exposure figure where that is what it reports); "results not disclosed" means the bank describes the system but no figure.

Each row quotes the bank's page on this site; the bank link opens the full record with every source. “Not disclosed” means the page records nothing on that point. Banks absent from the table have no evidence on this topic on their page.

Which banks lead on fraud AI specifically, and which on credit AI?

The two domains ranked separately by use cases in production, then described; ten each.

How is this ranking computed?

Score = 2 × fraud, AML/KYC, credit-underwriting and fair-lending use cases in production + 1 × those at other stages + 2 × disclosed result numbers about fraud, AML, approvals, losses or alerts (capped at two). Use cases are the ones tagged with those categories on the bank's page; the tag is applied when the page is written and validated against the use-case list. Ties: systems in production, then assets.

The second table ranks the two domains separately by use cases in production, then by use cases described, then assets.

What are the caveats?

Fraud AI is close to universal in large banks and often undisclosed in detail: a bank that names only its vendor's fraud model scores one use case, a bank that describes four systems scores four. Credit decisioning is under fair-lending and adverse-action rules, so banks describe it cautiously; several run AI in underwriting without saying so.

Result figures are on different bases (losses avoided, approval-rate lift, alerts reduced, payments screened) and cannot be compared across banks.

How does this differ from the AI index?

The AI index treats every use case alike; this page keeps only fraud, AML and credit use cases, weights production, and adds the disclosed result where there is one. The AI index and its method are on their own page; the technology leaders and generative AI guides are hand-written cross-bank tables with their own sourcing.