Which banks lead in AI for auto lending and auto finance?

Which banks have put AI or automation in auto lending on the record?

Every bank with an auto-finance use case, dated move or figure on its page that describes AI, models or automation: 9 of 140, the top 9 shown. Columns: the auto-finance use cases (with stage) or, where none is tagged, the dated moves; the first disclosed figure; the earliest dated evidence; and whether the bank itself calls it AI. Fair-lending and enforcement items are kept out of the score and listed in the second table.

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.

What fair-lending or regulatory record in auto lending is on each bank's page?

The latest auto-finance enforcement, settlement or fair-lending item on each ranked bank's page, plus any unranked bank with one. 4 banks carry such an item. None of these actions concerned AI; they concern dealer markups, servicing and underwriting rules that automated and AI decisioning in auto must now follow. "None recorded" means the bank's page has no such entry, not that the bank has a clean record.

How is this ranking computed?

Score = 2 × auto-finance use cases in production + 1 × auto-finance use cases at other stages + 1 per dated timeline move on auto finance (capped at 4) + 2 × disclosed auto-finance figures (capped at two). An entry counts when it mentions auto loans, auto finance, dealers, vehicles, indirect auto, floorplan or repossession and also describes AI, models, scoring, automated decisioning, fraud screening, pre-qualification or agents. Phrases that only look like auto finance (wealth 'Auto Commentary', contact-centre 'auto summary', small-business 'auto-decisioning', deposit 'auto-approval', home-and-auto insurance, broker-dealers) are removed before matching. Ties: assets, then name.

Only about 9 of the 140 profiled banks have auto-finance evidence on the record, so the table shows the top 9, not 30. "AI or machine learning stated" means at least one counted entry uses the bank's or the source's own AI, machine-learning or agent wording without a negation; everything else is labelled automation. Enforcement, settlement and fair-lending entries are listed in the second table and never add to the score.

What are the caveats?

Automation is not AI unless the bank says so. Automated credit decisioning, scorecards and dealer pre-qualification are on the record well before generative AI (Wells Fargo reported 59% of auto decisions automated in 2020); 3 of the 9 banks here describe only that kind of system, and the score treats a described automated system like any other use case. Read the last column before reading the rank as AI leadership.

Many profiled banks have small, run-off or no auto books and so nothing to disclose; others lend through dealers at scale but describe their decisioning only in filings. Banks describe auto decisioning cautiously because ECOA, Regulation B adverse-action rules and past dealer-markup cases apply; absence from this table is not evidence that a bank uses no models. Entries where auto finance is named only as a planned extension of another product are not counted.

How does this differ from the AI index?

The AI index scores AI disclosure across every business line. This page keeps only auto finance, counts automated decisioning and dealer tools alongside AI where the bank describes them, labels which banks actually say AI, and lists the auto fair-lending record separately. 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.