SYSTEM Cited by 1 source

Stripe Smart Disputes¶

Stripe Smart Disputes is Stripe's AI-powered dispute management product. It compiles and submits evidence on the merchant's behalf when a chargeback / dispute is raised by a cardholder, and — per the 2026-05-27 disclosure — develops a customised strategy per dispute rather than applying a uniform evidence template.

Capabilities at 2026-05-27¶

AI-recommended evidence¶

"Smart Disputes analyzes each dispute and surfaces AI-powered recommendations for specific evidence fields, such as tracking numbers or customer usage logs."

Disclosed gain: "Businesses that add our AI-recommended evidence through Smart Disputes are winning 3x more often than those that don't add any evidence." (Comparison is vs no evidence, not vs human-curated evidence.)

Evidence library¶

Evidence library is a merchant-curated document corpus — terms and conditions, return policies, service agreements — uploaded once and auto-selected per dispute:

"With the evidence library, you upload and store these documents once, and Smart Disputes automatically selects and includes them in your evidence packet based on the dispute's reason code, network requirements, and cardholder claims — no manual resubmission needed."

The architectural shape is the evidence-library with auto-selection pattern: per-document-type semantic routing over a static merchant-supplied corpus, gated by three orthogonal keys — reason code × network × cardholder claim. Selection mechanism (rule-based / learned / LLM-classified) undisclosed.

Position in Stripe's fraud / dispute stack¶

Smart Disputes is the post-dispute counterpart to systems/stripe-radar's pre-transaction signals:

Radar predicts likely-fraudulent transactions (see concepts/early-fraud-warning, concepts/fraudulent-dispute-prediction) and supports preemptive refund (see patterns/preemptive-refund-on-early-fraud-warning).
Smart Disputes activates after a dispute is filed, optimising the evidence packet for win rate.

The two pieces compose: Radar's fraudulent-dispute-likelihood prediction can feed Smart Disputes' evidence-gathering strategy before the dispute is even formally filed.

Operational numbers¶

3× dispute-win rate for merchants adding Smart Disputes' AI-recommended evidence vs adding no evidence.

Seen in¶

sources/2026-05-27-stripe-expanding-stripe-radar-to-protect-more-of-your-business — first canonical wiki disclosure; introduces AI-recommended evidence and evidence library.

systems/stripe-radar — companion fraud system; pre- vs post-dispute split.
concepts/evidence-library-for-disputes — the document corpus primitive.
concepts/fraudulent-dispute-prediction — Radar prediction feeding into Smart Disputes' strategy.
patterns/evidence-library-with-auto-selection — auto-selection pattern.
companies/stripe — vendor.