Africa’s largest lender, Standard Bank, says an ethical and responsible approach should be at the core of how businesses use artificial intelligence (AI).
AI investment has grown exponentially in recent years, driven by the rapid adoption and popularity of OpenAI’s ChatGPT since it was launched in November 2022.
Companies worldwide have felt the pressure to capitalise on the trend by employing AI-backed services or software platforms to either improve their own operations or create products.
Standard Bank makes the case that the advent of AI requires a fresh approach to how businesses, across sectors, engage with the technology internally and in the development of products that go out into the external market.
To do this, Standard Bank Corporate and Investment Banking (CIB) partnered with Stellenbosch University’s School for Data Science and Computational Thinking to develop a framework for ethical AI.
The project was aimed at proactively complying with evolving legal requirements, while keeping pace with emerging trends in AI.
“We decided to embark on this journey with the university because they have this deep specialisation in ethics,” says Heidi Taylor, legal specialist: digital technology at Standard Bank, in a recent episode of the Business Day Spotlight podcast — listen to it below.
“The part that really got us excited was the operationalisation of that framework and understanding how one can apply those ethics in the business-to-business space.”
Asking the right questions
Standard Bank CIB’s head of digital insights John Mukomberanwa emphasises the practical value of the research and how it can be applied to business decision-making.
“One of the first things that you have to think about when you’re implementing an AI system is asking the critical questions: what fundamentally changes by implementing a new AI system? Is it going to change the experience of customers? Is it going to help reduce losses? Is it going to help avoid risk? Is it going to help unlock new markets?”
For Mukomberanwa, these considerations form part of a broader process that begins in the discovery phase, where businesses identify the problem they are trying to solve with AI and bring the appropriate stakeholders into the discussion.
Addressing AI risks requires collaborative input from various specialisations, including security, IT, data, and legal teams. It cannot be viewed through a narrow, single-risk lens.
A major theme is that ethical considerations must be prioritised from the start of the design, development and deployment of such systems.
Applying ethics iteratively and continuously prevents it from becoming an overwhelming compliance hurdle that has to be retrofitted after the fact.
To translate research into practical product development, the ethical framework uses an array of about 120 guiding questions. These questions help businesses identify risks and potential blind spots, such as those relating to client needs and underrepresented groups, and assess the tangible value of an AI solution.
“[There are no] right or wrong answers. They are just questions where you need to actually apply your mind around the suitability [of AI] for your business,” Mukomberanwa says.
Taylor notes that while legal compliance is often black and white, ethics involves softer, subjective elements that require questioning what is right in a specific circumstance, rather than just what is strictly legal. It forces businesses to consider who is benefitting from the AI solution and whether any groups are being excluded.
Ethics as an enabler
Rather than acting as a roadblock, Standard Bank believes a responsible AI framework functions as an enabler.
“We see ethical considerations as an accelerator, in terms of actually sifting and filtering out things that may look nice on paper, things that may look nice as an idea, but that are actually not practically implementable,” says Mukomberanwa.
“It can be a game changer in terms of surfacing value for you early on, and actually making a decision on whether it is worthwhile to spend so much effort or money in terms of applying an AI solution, or maybe even prioritising what AI problems you need to work on.”
This ethics-by-design approach promotes transparency, accountability and human oversight, while providing guardrails to help mitigate risks as AI systems become a core part of the business.
Standard Bank has actively applied these principles to a corporate banking solution called Smart Nudge, an AI system that provides actionable insights to clients for risk management and liquidity opportunities.
For more insights, listen to the full conversation with Taylor and Mukomberanwa in the podcast above.
This article was sponsored by Standard Bank CIB.
