Sardine launched Sardine AI Labs, an applied research group focused on AI for financial crime detection.
Introducing Sardine AI Labs ↗ · repeat of an earlier article
What happened
- Sardine reports a held-out consumer card issuer fraud detection accuracy improvement of 68% (AUC-PR).
- Sardine reports a held-out business card issuer fraud detection accuracy improvement of 41% (AUC-PR).
- Sardine offers up to five research fellowships totaling $375,000 to independent researchers enrolled in US or Canadian universities.
- Applications for the fellowship close December 13, 2026.
Why it matters to Signicat
- Sardine positions foundation-model features as an add-on to existing fraud models in card issuing, a fraud-prevention direction adjacent to Signicat's risk and AML screening offering.
- Sardine's 68% and 41% figures are vendor-reported offline results on held-out issuers, not independently verified; no benchmark method is independently established.
Evidence from the source (6 quotes)
- “| Consumer card issuer | 68% | ” source ↗
- “| Business card issuer | 41% | ” source ↗
- “Sardine AI Labs is offering up to five research fellowships, totaling $375,000, to independent researchers currently enrolled in universities in the US or Canada” source ↗
- “Applications are open through December 13, 2026.” source ↗
- “give existing risk models a richer representation of customer behavior rather than replace the entire risk stack.” source ↗
- “The initial results:” source ↗
Fraud deepfake injectionAML KYB screeningResearch reportProduct launchEvent webinarGlobal
5 of 5 facts verified against the source · confidence medium · claude-haiku-5-5