SBS launches sovereign AI platform for core banking

SBS unveiled what it calls the first sovereign AI platform built for generative and agentic AI in core banking, with data kept inside a bank’s own environment. The launch targets banks facing strict data residency and regulatory rules, and the platform is already running in production with Scottish Building Society.

Why it matters: - Banks are under pressure to use AI to cut costs, speed decisions and compete with digital-first challengers. - Many banks cannot send sensitive customer and transaction data to external AI services because of GDPR and country-specific data residency rules. - SBS is aiming to close that gap by letting financial institutions run generative and agentic AI without moving data outside their own controlled environment.

What happened: - SBS launched SBS AI Foundation, a sovereign AI platform built for generative and agentic AI in core banking. - The platform is powered by NVIDIA and is available to the more than 1,500 financial institutions SBS serves worldwide. - The system is already live in production with Scottish Building Society in the UK on Amazon Web Services. - SBS said the platform can run on a private cloud, a managed cloud or fully on-premises.

The details: - The platform uses NVIDIA infrastructure, including NIM microservices and NVIDIA NeMo Guardrails. - SBS said the setup keeps every response within a bank’s regulatory boundaries. - The platform pulls data from across SBS banking systems, including core banking, lending, digital banking and regulatory reporting. - SBS said the solution gives generative AI and autonomous agents immediate access to that data. - In production with Scottish Building Society, the platform is already delivering about 30% to 40% faster response times and more consistent outputs. - SBS said the system preserves explainability and a full audit trail for regulators. - The platform supports generative AI grounded in an institution’s own data for onboarding, engagement, product matching, personalization and product design. - It also includes natural language intelligence that turns complex data queries into insights and reduces multi-day reporting cycles. - A third use case is sovereign risk intelligence that surfaces risk and compliance exposure while generating a complete audit trail. - SBS said the platform will soon include agentic AI for multi-step banking workflows with human oversight. - The launch marks what SBS called the first deployment of NVIDIA’s full AI stack, including NIM, NeMo Guardrails and open Nemotron models, inside a live core banking system. - SBS said the deployment spans markets with strict data residency rules, including GDPR-regulated markets in Europe and data-resident markets in Africa.

Between the lines: - The launch is as much about compliance architecture as it is about AI capability. - Banks have wanted AI gains for years, but many have lacked infrastructure that can support large-scale AI inside regulated environments. - SBS is positioning governance and auditability as the main reason banks can adopt AI faster. - NVIDIA is gaining a deeper foothold in financial services by moving its stack into regulated core banking workloads.

What's next: - SBS plans to expand the platform’s agentic AI capabilities. - The company expects the new infrastructure to help more financial institutions adopt AI efficiencies while keeping data under local control. - Early customer results will likely shape how quickly SBS rolls the platform out across its banking base.

The bottom line: - SBS is betting that banks will adopt AI faster if they can keep data, controls and audit trails inside their own walls.