Artificial intelligence (AI) is forcing a shift in the way businesses operate and is the foundation to support the construct of an agentic-powered enterprise.
Global technology and consulting firm IBM warns though that governance as an afterthought and a siloed approach to AI adoption remain challenges.
Speaking at the ITWeb Chartered CIO Conference 2026, hosted recently in Johannesburg, Sabine Holl, VP sales engineering and CTO, IBM MEA, said organisations are looking to leverage agentic AI to build business at scale, but are struggling to not only keep up with the technology, but to do so with effective governance in place.
The challenge comes as South African organisations accelerate AI adoption. According to the IBM Institute for Business Value 2026 CEO Study, 83% of surveyed South African CEOs say they are actively embedding AI across multiple workflows to improve business effectiveness, while 67% have already appointed a chief AI officer, signalling that AI is becoming a core part of enterprise operating models rather than a standalone technology initiative.
Holl said AI has become more commoditised and is enabling faster adoption, but with that comes the challenge to control multiple agents.
IBM says it is positioned to strengthen and support the requisite infrastructure to manage AI projects, delivering value in agentic app-driven initiatives to increase productivity, procurement and sales.
The company suggests there are strategic imperatives for building the AI-native enterprise, including real-time data infused in agentic apps, resulting in better insight for clients.
“In terms of running IT operations and securing the fastest return on investment in AI, we don’t lack tools – it’s a matter of bringing all the data together and being able to draw the right insights from the data,” Holl continued, adding that everything must be underpinned by security.
Holl emphasised an increasingly popular AI use case – the application of the technology across the software development lifecycle, and specifically the contribution of the company’s agentic AI coding partner, IBM Bob.
“This is about using AI to build control across all tools, making the right decisions and at speed, as well as minimising cost exposure to fix problems,” Holl added.
Software developers are using Bob and its automation capability to leverage IBM infrastructure.
Moving to AI
“We are talking about AI-augmented operations and an increase in productivity,” said Holl.
Scattered data and siloed AI add complexity to AI operations, and is one of the reasons why IBM developed the IBM Concert Platform, an agentic operations platform that connects data, context and actions across the environment.
The platform is a solution designed to empower organisations to move from insight into action, says IBM.
The company explains that across hybrid environments, teams depend on a patchwork of tools for observability, optimisation, security and operations. While each platform generates valuable signals, they operate without a common context. This creates an execution gap − the widening disconnect between what teams know and their ability to act on it.
As a result, incident resolution remains slow, cost predictability becomes increasingly difficult, risks accumulate unnoticed, and teams spend more time piecing together information than driving meaningful action.
This disconnect is what IBM Concert Platform is designed to address, by helping to cement coordinated agentic operations.
Holl added that as organisations transform and adopt agentic AI, there is an “explosion of apps”. According to IBM, there will be one billion new apps by 2028.
IBM has readied its solution portfolio to help organisations manage increasing levels of complexity in operation.
Security framework
Holl said amid this rush to onboard agentic AI and the need to bolster productivity at scale, the issue of security warrants closer and more deliberate attention.
The level of complexity surrounding AI continues to increase, particularly as applications become more agentic and introduce additional security touchpoints. At the same time, sovereign AI has become a genuine boardroom discussion.
While sovereignty was not always taken as seriously as it should have been, organisations are now recognising its importance in hybrid application environments that span public cloud, private cloud and on-premises infrastructure. Data sovereignty raises critical questions around who controls data, digital infrastructure and AI compute power.
Holl added that from a cloud adoption perspective, IBM has learned from customers in the Middle East that sovereignty is fundamentally about control rather than location. However, from a regulatory perspective, organisations must broaden their thinking and carefully consider multiple dimensions of digital sovereignty, including geographic requirements.
Backup and recovery plans are not simply boxes to tick, but rather a fundamental and critical component of any sovereignty strategy. Through its sovereignty workshops, IBM helps organisations understand these digital sovereignty dimensions and develop appropriate approaches, Holl said.
IBM is enabling organisations to build sovereign environments through innovations such as an AI Factory in a Box, allowing enterprises to run AI capabilities on-premises within their own organisations. Combined with cloud-type services tailored for different entities, these technologies help clients improve resilience, compliance and sovereignty.
IBM Sovereign Core redefines how sovereignty is implemented, governed and proven, with building blocks that run on top of OpenShift.
Key takeaways from the Chartered CIO Conference are that IBM continues to build new technologies, remains a strong force in quantum computing, and has developed quantum-safe encryption methodologies designed to protect organisations in the future, it says.
Share
