Ivanti’s 2026 AI Maturity Report surveyed 3,900 employees across six countries and landed on a finding that should give every IT leader pause: the gap between how fast organisations are deploying AI and how well they are governing it is widening, not closing.
More than half of organisations (56%) are now deploying AI broadly across multiple workflows, or at business-critical scale. Only 2% report no AI use at all. That is not a cautious, experimental picture. That is mainstream enterprise adoption, moving fast, with real operational consequences.
The results are real: 45% of IT workers say AI makes their work faster and better, and 63% say they spend less time on repetitive tasks. But underneath those numbers sits a governance problem that mature organisations are managing well, and less mature organisations are largely ignoring.
The Accountability Gap Nobody Is Talking About
Here is the statistic that stops the conversation: 85% of IT professionals claim there is a named, accountable owner for every AI agent and workflow in their organisation. But only 42% say that accountability is actually clear.
That is a 43-point gap between what organisations say they have and what is true in practice. And it matters, because when accountability is unclear and policies are inconsistently followed, AI operates from fragmented, unverified data rather than a trusted system of record.
The picture gets more uncomfortable when you look at policy compliance. Among companies that have AI policies in place, only 24% of employees say those policies are followed very consistently in day-to-day work. Another pressure point: organisational leaders are nearly twice as likely to keep their AI use secret compared to other employees (42% vs 23%). Of those hiding their usage, 52% say they do so for a “secret advantage.”
Governance is now the most commonly cited barrier to faster AI deployment, according to the report, with 27% of IT professionals identifying governance, security or compliance concerns as the biggest obstacle. That exceeds skills shortages (20%), technology limitations (17%) and data challenges (14%).
What AI Is Actually Doing in IT Operations
Within IT service management and endpoint management, AI is already embedded at meaningful scale. In ITSM, the most common applications are virtual agent and chatbot support (58%), ticket classification and routing (56%) and automated ticket resolution (51%).
In endpoint management, AI is handling anomaly detection (57%), device vulnerability identification (55%) and patch prioritisation (47%), with far greater automation expected within the next 24 months.
Agentic AI is no longer an experiment: 57% of IT organisations are using it for at least several important workflows, including 17% who rely on it for extensive end-to-end workflows. Deployment is concentrated in Level 1 support (61%), network and infrastructure operations (59%) and Level 2 specialist support (57%), precisely because these are the highest-volume, most repetitive functions in IT.
The trajectory is steep: 46% of IT organisations expect AI to automate nearly half of their operations within 18 months.
The Maturity Divide Is Real, and It Is Measurable
The most telling finding in the report is the performance gap between organisations at different stages of AI maturity. At organisations where AI is scaled or business-critical, 54% of IT professionals say AI makes their work both faster and better. At early-experimentation organisations, that figure is 24%.
The gap in proactive issue detection is even wider. At early-experimentation organisations, 43% of IT professionals say AI frequently helps their team detect issues before end users are aware. At scaled organisations, that number is 89%.
Mature AI organisations are not just doing more of the same things. They are operating at a categorically different level of IT performance and resilience. The report’s data shows this across time to resolution, customer satisfaction scores and cost savings, with scaled organisations measuring all three at roughly double the rate of early-stage peers.
Advanced individual AI users save an average of six hours per week, double the three hours saved by the least mature users. And intent to upskill tracks directly with AI maturity: 86% of the most advanced users plan to update their skills in response to AI’s impact, compared to just 37% at the basic level.
What This Means for IT Leaders
The report draws a clear line between organisations that treat governance as an afterthought and those that embed it from the start. The difference is structural, not philosophical.
Scaled organisations that have closed the governance gap share a common approach: every AI agent and workflow has a named owner, escalation paths are defined before they are needed, and accountability is automated into policy and practice rather than declared in a document that nobody reads.
Three specific imperatives come through clearly in the research:
- Build governance in, not on. Accountability needs to be structural, embedded in the platform and enforced automatically, not bolted on after the fact.
- Redesign roles, not just workflows. More than a third (37%) of organisations report that at least a few IT roles or teams have been significantly reshaped by AI. The organisations getting the most from AI are rethinking what their teams do, not just augmenting the same job descriptions with AI tools.
- Make upskilling systematic. The gap in AI literacy between IT professionals and the broader workforce is a source of friction. IT pros rate their own AI literacy as high or very high at nearly double the rate of office workers (62% vs 27%). Closing that gap requires intentional, organisation-wide education programmes, not ad-hoc learning.
The Window for Measured Action Is Narrowing
The question is no longer whether AI will transform IT operations in South African enterprises. The global data suggests it already is, and the trajectory points firmly upward. The real question is whether IT leaders will shape that transformation deliberately, by building the governance, skills and structural accountability to match their deployment pace, or be pulled along by it.
The maturity divide in Ivanti‘s research is not a gap between organisations that chose AI and those that didn’t. It is a gap between organisations that invested in the foundations and those that skipped them. The good news is that the data shows governance improves dramatically with maturity: 69% of scaled organisations report comprehensive governance in place, compared to just 15% at early experimentation stage.
Getting there requires a deliberate decision to treat AI governance as infrastructure, not afterthought.
Read the full 2026 AI Maturity Report from Ivanti to explore the complete data set, including country-level breakdowns and the executive summary with editable slides.
