SBS

Deposits, payments, cards and lending ran for decades before any model arrived and would continue with every model removed. The learned line is named: machine learning financial crime detection and the SBS AI Foundation generative layer announced in July 2026.

Guarded workflows are promised for its agents, and collections prioritization is described as automated, yet no threshold, review requirement or exception route is set out. Advisory by design, the products give relationship managers a view of customer needs, but nothing says what a person must approve before a score or collections action reaches a customer.

Explainable decision making is promised without a described mechanism, and no accuracy figure, validation method or drift disclosure is published for any model. Nothing says how a bank would validate what it deploys.

Santander, BNP Paribas and HSBC are among many named institutions, and Bank11 extended its partnership by eight years, but no result is attached to any of them. Its one performance figure, financial crime false positives cut by up to 85 percent, belongs to no institution.

Institutions are promised control of their data and AI, with AI Foundation data said to remain governed within each institution's environment. Yet on what SBS calls a shared data and AI platform, nothing says whether customer data trains any model or is pooled across institutions.

The June 26, 2026 data processing addendum makes SBS processor on the institution's instructions, names subprocessors in annexes with fifteen days' notice of changes and a right to object, and deletes or returns data at contract end. Processing runs across eight countries from France to the United Arab Emirates under standard contractual clauses, with no retention periods given.

Contractual security plans are published openly, versioned and dated, and the SaaS plan states ISO/IEC 27001:2022 certification scoped to the cloud native offering, with an extension planned. Clients with an audit right receive audit reports, though SBS points to AWS Artifact for SOC reports rather than naming one of its own.

Inside the listed group 74Software, SBS states no license or supervised standing of its own. Regulatory references concern the buyer's position, from open banking rules to regulatory reporting delivered as a product.

AI supported scoring and risk assessment run inside loan origination, and many of the lenders sit in the European Union, where creditworthiness assessment of natural persons is high risk under the AI Act. Neither fairness testing, disparate impact analysis nor adverse action support is published, and no position on that regulation appears.

Audit rights and full liability for subprocessors' data protection obligations go to the institution under the addendum, but SBS publishes no commitment behind a model output. A false sanctions hit freezes a real customer's payment, and that customer complains to the institution with no route to SBS described.

Critical suppliers are named in the SaaS security plan: Amazon Web Services for cloud, MongoDB Atlas for the database and IBM Safer Payments for fraud detection. SBS AI Foundation names no model provider, family or version.

As supplier of the system of record, SBS lets lending, asset finance and compliance attach natively, with composable components sharing that spine. Data from the products a bank already runs feeds the AI layer directly, and open banking support supplies the outward interface.

Separate security plans cover software as a service, with client data stored in European Union data centers, a managed service and on premises under a maintenance contract, and SBS AI can run on public cloud, private cloud or on premises. No regions are named, and the SaaS location is European only.

Security plans, the data processing addendum and SaaS agreement documents are all linked openly from the footer, yet no pricing of any kind appears. Whether SBS AI Foundation is licensed separately from the SaaS, managed service or on premises shapes is not stated.

More than fifteen hundred institutions in eighty countries span distinct buyer types, from a British building society to African banking groups such as KCB and captive finance arms such as Toyota Financial Services. Product lines run retail, corporate and dealer floorplan financing.