Standard Bank has partnered with Stellenbosch University's School for Data Science and Computational Thinking to produce a framework for responsible AI in business. Called the Standard Bank Responsible AI Approach, it sets parameters for developing, deploying and using AI, and covers human oversight and accountability across a system's life cycle.

The timing is the interesting part. Three days ago the FSCA said it would wait for the Financial Stability Board's global findings before setting any local AI rules, and that its approach would be principles rather than rules because the field moves too fast. So the country's largest bank by assets has written its own framework into that gap, with a university supplying the ethics and computational expertise.

John Mukomberanwa, head of digital insights at Standard Bank Corporate and Investment Banking, says AI brings real gains in efficiency and client experience but introduces uncertainty through its complexity, and that the bank went to an academic partner deliberately to pair industry knowledge with data science and ethics.

Professor Kanshukan Rajaratnam, who directs the school, says the framework can be applied by any business and should be treated as a living document that changes as the technology does.

What has not been published is the framework itself. No document, link or summary of its contents has been released, so what it actually requires of a team building an AI system is not yet public.

ITWeb reported the announcement and the bank's argument that scrutiny of business AI has grown on ethical and human rights grounds, covering bias and discrimination, transparency, accountability and privacy, and that these concerns weigh heaviest in financial services where decisions affect livelihoods directly.

Statista estimates 30% of B2B companies were actively implementing AI across their organisations in 2025, up from 23% in 2024. Stellenbosch's School for Data Science and Computational Thinking is the academic partner on the work.

This is what happens when a regulator steps back. The FSCA is waiting for Basel and has chosen principles over rules, which leaves a vacuum, and the largest institutions fill vacuums. A framework written by Standard Bank will shape what its suppliers, partners and corporate clients are expected to demonstrate, long before anything is gazetted.

That is not a criticism, because somebody has to go first and pairing it with a university is more rigour than most would bother with. But if you sell software or services into South African financial institutions, the practical effect is that a bank's internal document may become your compliance requirement.

Ask for it: the bank says it applies to any business and the professor says it has wider societal relevance, so there is no obvious reason it should stay unpublished.

You might also like our piece on the FSCA waiting for global guidance on AI, why SA enterprise AI adoption keeps outrunning its own strategy, and how AI financial crime in SA outpaces compliance.

Get more SA tech and business news and subscribe to The Open Letter.