Nearly half of enterprises investing heavily in AI are struggling to see meaningful returns, according to Kyndryl’s own Readiness Report.
This finding is echoed by an IBM study showing just 25% of AI initiatives deliver expected ROI, and only 16% have scaled enterprise-wide.
The ‘why isn’t it working?’ question has been a fixture of analyst reports and CIO conversations for years. Kyndryl is now building a business around answering it.
Last month, the IT infrastructure giant launched Agentic Service Management, a framework combining a maturity model, structured assessments, and implementation blueprints designed to help enterprises move from traditional service operations to autonomous, intelligent workflows.
The argument behind it is that organizations neck-deep in AI investment and thin on AI returns need to hear, as Kris Lovejoy, Global Head of Strategy at Kyndryl, explained:
“Most enterprise environments were built for people running tickets and tools, not for fleets of autonomous agents executing tasks across hybrid and multi-cloud estates – and this mismatch is limiting AI from moving out of pilots to outcomes.”
This might be blunt, but it is hard to argue with. For a lot of service operations leaders, it puts a name to a problem they’ve been circling for months.
A Maturity Model for the Agentic Era
The Agentic Service Management offering is delivered through Kyndryl Consult and starts with an assessment of where an organization actually stands.
That means evaluating its current state across service management, AI governance, security, and operations, then benchmarking those capabilities against emerging standards – including ISO 42001, the AI management standard that most enterprises haven’t meaningfully engaged with yet.
From there, Kyndryl delivers a tailored gap analysis and a phased roadmap. Also available as a standalone service is Kyndryl Agentic AI Digital Trust, which provides a security-first framework for governing agentic AI deployments across hybrid and multi-cloud environments.
For regulated industries (financial services, healthcare, public sector) this is arguably the more urgent piece. An AI agent operating outside defined boundaries in those environments can become a significant compliance issue.
Rather than leading with the technology, Kyndryl is leading with organizational readiness. Plenty of vendors are selling AI agents.
Fewer are addressing what the business itself needs to look like before those agents can operate reliably at scale.
As Lovejoy put it:
“You can’t scale agentic workflows on top of operating models that were designed for manual work.”
Kyndryl Is Applying This to Its Own Operations
Kyndryl isn’t just talking the talk; it’s walking the walk. The company is applying Agentic Service Management to its own internal service delivery operations through Kyndryl Bridge, its open integration platform.
That existing automation foundation already runs nearly 200 million automations per month across more than 8,000 certified playbooks.
That base means the maturity model is grounded in running mission-critical infrastructure at enterprise scale, not assembled in a strategy document. For potential customers doing their due diligence, that distinction could be crucial.
The Bigger Picture
Zoom out, and the context is a market shifting from AI strategy to AI execution. The heavy investment cycle of the last two years has left a lot of organizations sitting on a portfolio of pilots and a stubbornly flat ROI line.
Kyndryl’s answer is that governance, workflows, and controls are still rooted in the pre-AI era – and that the operating model, not the technology, is the real bottleneck.
For a managed services company, that positioning helps to shift the conversation from platform selection to operational transformation, which is firmly in Kyndryl’s territory.
There’s a direct CX angle here, too. Agentic service management doesn’t stop at IT operations. As AI agents take on more customer-facing work – such as case resolution, proactive outreach, self-service escalation – the same governance and oversight principles apply.
The organizations building those internal controls now will be in a stronger position when customer-facing deployments start to scale, and regulators start to ask questions.
The honest caveat is that maturity models have a long history of becoming shelf documents. Kyndryl has correctly identified the problem and built a structured way to attack it.
Whether that translates into measurable outcomes for customers – or stops at a gap analysis and a roadmap – is the question enterprises should be pushing hard on before they sign up.
