Standard Bank, Africa's largest lender, is moving experimental AI tools into production across product design, customer operations, and internal workflows. The shift signals a broader industry push to derive measurable value from generative AI rather than treating it as a technology showcase.
Andrew van der Hoven, chief digital and product officer for Personal and Private Banking at Standard Bank, said the bank is now using AI to accelerate how teams build products and resolve customer friction. The approach prioritizes specific business problems over technology-led deployments.
"You have to start with the client problem ... Otherwise, what can happen is that the technology can potentially start looking like the solution for every single problem and it might not be," he said in a recent podcast interview.
Faster prototyping and sharper product decisions
Generative AI is cutting product development cycles dramatically. Van der Hoven said the bank can now develop solutions between 50% and 70% faster, moving from concept to customer feedback in compressed timeframes.
Using AI, product teams can prototype hundreds of variations of a potential product and test customer preferences before committing to development. This replaces the older model of relying on expert judgment, building a finished product, and only discovering whether it worked after launch. The change directly affects how product managers allocate resources and validate ideas.
For professionals working in AI for Product Development, the approach illustrates a practical shift: AI is not generating final products autonomously but compressing the research and prototyping phases that traditionally consumed months.
Customer operations shift to natural language
Standard Bank has replaced traditional interactive voice response menus with natural language processing, allowing customers to explain needs in their own words. Routine queries are increasingly automated, freeing human bankers for complex interactions.
"About 65% of everything that our clients ask us [on our digital platforms] is actually resolved with those AI technologies," Van der Hoven said. The system transfers customers to human team members when needed, without forcing them through rigid menu trees.
The bank built an internal platform around a large language model to give employees context-aware AI tools. More than 20,000 staff members actively use the system. Van der Hoven described one tool that grew from fewer than 50 users to more than 1,000 "overnight" after a July conference where 1,200 team members received hands-on training.
Data and trust as the durable advantage
Van der Hoven cautioned that access to powerful AI models alone will not create lasting competitive separation. Those models are becoming commoditized.
"The durable advantage really lies in Standard Bank's trusted data and our deep relationships with our clients and how we integrate that with our culture to deliver these brilliant solutions," he said. The bank builds privacy, legal, and security considerations into AI solutions from the start, with human oversight treated as fundamental rather than supplementary.
He acknowledged that generative AI presents distinct challenges for financial institutions because its outputs are inherently probabilistic. For teams working with Generative AI and LLM Courses, the bank's emphasis on trusted data as the moat - rather than the model itself - reflects a maturing understanding of where competitive advantage sits in enterprise AI deployments.
Why this matters for product development professionals
Standard Bank's experience offers a concrete benchmark: a 50% to 70% reduction in development cycle time is achievable when AI is applied to prototyping and customer validation, not just code generation. The takeaway for product teams is that the fastest gains may come from using AI to test more ideas earlier - replacing expert guesswork with rapid customer feedback - rather than trying to automate the entire build process.
Your membership also unlocks:
