For more than a decade, automation inside banks has been used discreetly and on a distributed basis, principally for task automation. It handled bits of processes, typically within business silos.
That limited, incremental approach made sense when robotic process automation (RPA) was deterministic, fragile and reliant on stable, rules-based flows.
However, the world has shifted. The combination of artificial intelligence and automation has now reached a point at which we can address enterprise issues within an organisation.
Banks are making huge leaps in AI adoption, but there is much more work to be done.
The gap between what the technology can already do and how organisations still imagine their businesses is where iqbusiness sees the biggest challenges, and the biggest opportunities for banks as they shift their thinking to match the evolving capabilities.
The problem isn’t technology
Every major technological shift comes with a lag between what you can do with the technology and the imagination of what you could do with businesses. The same thing is happening now with AI – the tech is ready, but our experience shows us that many banks are still playing catch up.
Most of today’s banking operations still reflect the century-old paradigm of scientific management thinking. This way of operating essentially deals with how we organise people for work.
Everything we take for granted, hierarchies, spans of control, working hours, handoffs, inbox queues, maker-checker steps and more comes from organising human beings. However, all these attributes are about organising people, not designing an optimal operational model for the 21st century.
AI and automation don’t operate within those constraints. They operate on access, logic, orchestration and speed.
The new dispensation has the potential to run roughshod across every bit of that. This is the fundamental challenge that faces bankers today: organisations are applying 20th century mental models to 21st century capabilities. Successful banks will be those that can bridge the gap.
The biggest internal barriers banks face
Sensitively balancing the competing priorities of banking teams’ objectives is vital to get right. A digital worker is an operations asset which immediately pulls the IT, ops and executive teams into conflict. A lot of the time, IT might prefer to replace or enhance IT systems, while the chief operating officer seeks operating model transformation to achieve business goals.
These are very real, very human and absolutely fundamental issues to identify and solve from the outset. While the business sees immediate possibilities for customer experience, cost optimisation, compliance improvements and faster turnaround times, IT is often vested in practicalities like:
- Protecting legacy systems that are tried, tested and reliable;
- The belief that AI and automation are tactical fixes, not systemic solutions;
- Concerns about security, access and credentials;
- Viewing bots and agents as “non-standard”, or shadow interfaces; and
- Fear of losing control of the technology stack.
Our experience is that it can never be one without the other, and the right interventions that ultimately build better human relationships can make or break a bank’s success.
The next barrier is using automation in isolation rather than at scale. Despite a decade of RPA, we find that people still buy a bot for a particular task within a particular process within a particular area, and it’s only used for maybe 50% of the time.
This limiting (and expensive) approach is a consequence of mirroring the organisational structure rather than the operational need to automate customer journeys.
It also prevents the economies of scale that make AI and automation transformative. A vital shift from the former to the latter is critical to unlock these capabilities and ensure that the underlying banking business continues to evolve and grow.
The next stumbling block is legacy processes that assume humans are doing all the work. Banking processes were historically designed around people. That means they are linear, slow, sequential, full of handoffs and dependent on memory and judgment. They are also bottlenecked by working hours and delayed by queues, inboxes or capacity constraints.
But we know already that AI and automation shatter these assumptions. They don’t work linearly or wait for inbox refreshes. They don’t stop at 5pm, and they don’t need a queue.
In our local and international work, we frequently find that banks are still designing workflows that assume all of those constraints are unavoidable.
Critical capabilities that are missing internally
One is the ability to holistically design for people, bots and AI agents together. Future operating models aren’t human-centric or machine-centric. Rather, they are designed from the start around three categories of workers:
- People (handling exceptions, judgment, relationships)
- Digital workers or software robots (handling repetitive, rules-based work)
- AI agents (handling ambiguity, non-deterministic flow, reasoning, back-and-forth tasks)
Most banks don’t yet design processes this way.
The second missing capability is AI-powered orchestration, which is one of the key differentiators of iqbusiness. Our teams use AI-powered orchestration to distribute work between people, bots and AI agents, and to manage the flow of a task from the beginning to the outcome.
What is also often missing is a centralised digital operations control room. At scale, automation becomes an operational estate, not isolated bots. And this is where iqbusiness believes the heart of the future of banking lies: the creation of a centralised control room for the enterprise (like a Nasa control room) handling variable demand 24/7, built on the skills, experience and infrastructure of a multiplicity of experts.
When success takes minutes, not weeks
In a recent project for an African bank, corporate onboarding had historically taken weeks. iqbusiness re-engineered it using AI agents for verifications and to write approval briefs, digital workers for extraction, classification and running tasks, and AI orchestration to coordinate the confusing space between request and resolution.
Due to these changes, corporate onboarding now happens in under 20 minutes. This worked because the process was redesigned end-to-end and run in a swirl, rather than attempting to automate step-by-step.
A hugely important part of this redesign was to remove some key organisational boundaries that can offer comfort on paper, but cause havoc in practice. Because AI can effectively capture all the data in one burst across any system it has access to, old structures like maker-checkers slow things down and don’t actually act as risk limitations as intended. Here, the AI produces a complete, verified pack and a human being handles exceptions and the final decision.
Reusing agents across the enterprise is another benefit. When we build a verifications agent, it can do the verifications for any product in a banking suite. This is the opposite of today’s siloed “one bot per use case” mindset and is enormously transformative when correctly designed and applied.
What banks can do next
There are four simple steps that banks should be taking right now:
- Identify what people should be doing and start shifting mindsets from people-centric processes to mixed-worker operating models. Start by asking: “What do we want people to do?” And let automation handle the rest.
- Build the orchestration layer. This is the real intelligence: routing work, managing flow, handling escalations, and coordinating the digital workforce.
- Create the digital operations control room. Run the digital estate like a 24/7 workforce. Customer work by day, internal work by night. Importantly, design processes around outcomes and not organisational charts. The AI doesn’t care which department owns which step.
- Use AI agents for ambiguity, reasoning and back-and-forth problem-solving. This is where the biggest time savings and customer improvements live.
Real momentum will be found in banks that start in the right place, wrap systems with AI to speed turnarounds and generate new value from their transactional data.
These are the banks that will shake off constant backlog-management and find the sweet spot between AI, automation and people, interfacing through APIs to current systems. iqbusiness helps banks make the shift and get speed to value. We do so with experience gained from real-life references, scarce AI and automation-related skills, practical delivery know-how and an understanding of how agentic automation improves operations with a clarity that is difficult to generate quickly internally.
AI is here and it’s time to use it with purpose.
About iqbusiness
iqbusiness is Africa’s future-focused management and digital growth enabler, founded on over 26 years’ experience. Led by some of the continent’s best thinkers and doers, our purpose is simple: to grow people, business and Africa as one.
Established in South Africa in 1998, iqbusiness integrated into Reunert ICT in 2023 and merged with +OneX in 2024. Together, we offer a full-service suite of digital services, managed services, digital consulting and management consulting across the continent.
Our scale and unique capabilities mean we unlock exponential value and global growth for our clients. Renowned for our tenacious Geshido energy, and determination to deliver projects of excellence and meaning, we are a proud level 1 B-BEE contributor and hold B-Corp, conscious companies and top employer status. For more information, visit www.iqbusiness.net.
- The author, Steve Burke, is the executive for agentic automation at iqbusiness
- This promoted content was paid for by the party concerned
