SBS Puts Data Control at the Center of Its Banking AI Launch

The banking software provider says its NVIDIA-powered platform can run across private cloud, managed cloud and on-premises environments, with one deployment already live

For banks weighing generative AI against regulatory obligations, the central infrastructure question is not simply what a model can answer, but where the underlying information resides. SBS is positioning its new AI Foundation around that distinction. Announced October 13, 2026, the platform is intended to let financial institutions use AI while keeping data within their chosen institutional environment, according to the company.

SBS describes AI Foundation as sovereign AI for banking, with deployment options spanning private cloud, managed cloud and on-premises infrastructure. Those options matter because banks do not all operate under identical hosting, security or data-residency requirements. The announcement establishes deployment flexibility as a product proposition; it does not independently demonstrate that every configuration satisfies every institution’s legal, supervisory or internal governance obligations.

The platform is powered by NVIDIA NIM microservices, NeMo Guardrails and Nemotron, SBS said. CEO Eric Bierry framed the offering around control and accountability: “Banks want AI they can trust, explain and run wherever their data must live. By building with NVIDIA, we can give our customers that assurance without compromising on governance.” That assurance is the company’s position, rather than an independent certification of the platform.

AI Foundation is designed to draw on core banking, lending, digital banking and regulatory data. Its announced capabilities include generative AI grounded in an institution’s information, a natural-language co-pilot and sovereign risk and compliance intelligence. Bringing those data domains into a common AI layer could make the platform relevant to multiple banking functions, but the release does not establish the extent of adoption across those functions.

SBS identifies Scottish Building Society as an existing live deployment running on AWS. That is the specific production example supplied in the announcement, not evidence of a broader rollout across SBS customers. It also illustrates that the company’s sovereignty proposition is not limited to infrastructure physically located on bank premises: the stated deployment choices include cloud environments as well as on-premises systems.

For that deployment, SBS claims response times are 30% to 40% faster, alongside consistent outputs, explainability and an audit trail. The performance figures are issuer claims, not independently verified results. The supplied announcement does not provide enough detail to evaluate the comparison baseline or test conditions. A buyer would therefore need to distinguish the advertised improvement from performance under its own workloads and governance requirements.

Explainability and an audit trail are consequential features in financial services, where institutions must be able to examine how systems are used and how decisions are supported. Still, the presence of those capabilities does not itself establish regulatory compliance. For procurement teams, the practical assessment would concern what is recorded, who can review it and how those records fit existing controls, rather than treating the product description as a compliance determination.

The company says agentic workflows are planned for a later stage. That distinction separates the current announcement from a claim that autonomous workflows are already operating across customer banks. Institutions considering the platform can assess the announced assistance and intelligence features without assuming that the future agentic capabilities are available today. The release supplies no basis for treating that planned expansion as a completed deployment.

SBS says more than 1,500 financial institutions in 80 countries use its products, providing context for its potential distribution reach rather than a customer count for AI Foundation. The business case now rests on translating an existing banking software footprint into governed AI adoption. Scottish Building Society supplies an initial operating reference; wider uptake, comparable performance and institution-specific governance remain questions beyond the evidence presented in the launch.

Source: SBS Official site