Kyndryl report: AI is in 57% of enterprise core processes, but workforce readiness is falling behind

Fifty-seven percent of enterprises now have AI embedded in core business processes or deployed broadly across their operations, up from 35% a year ago. That is the headline number from Kyndryl's second annual People Readiness Report, which surveyed 1,100 senior business and technology leaders across eight countries. The finding confirms that enterprise AI adoption has moved well past the pilot stage. The problem is what's not keeping pace.

Only 32% of those organizations have achieved at least one of their two primary AI objectives. Just 11% have hit both. Rapid deployment, it turns out, does not automatically produce business results.

Workforce preparedness is moving in the wrong direction

Even as AI investment accelerates, organizational readiness to absorb it is declining. Only 23% of business leaders now believe their workforce is fully prepared for AI, a six-point drop from 2025, according to the Kyndryl report. Seventy-nine percent agreed that the pace of AI development will outpace their organization's ability to adapt its workforce, governance structures, and operating models.

The skills pipeline is tightening, too. Fifty-two percent of respondents said finding employees with the right AI competencies has become more difficult over the past year. Only one-third have fully implemented training programs designed to prepare staff to work alongside AI tools.

Kyndryl's CIO Kim Basile noted in the report that organizations seeing the strongest outcomes are investing in their people alongside their technology, rethinking roles, funding upskilling, and actively managing employee transitions through change.

A 9% cohort shows what good looks like

Kyndryl identifies a group it calls Pacesetters, roughly 9% of respondents. These organizations share a common approach: they redesign roles around AI capabilities, build structured change management programs, and establish governance guardrails before scaling deployments. The results are measurable. Pacesetters are approximately twice as likely as their peers to have fully implemented AI governance across every dimension the report tracks. They are also 1.5 times more likely to report AI-driven revenue growth and 1.6 times more likely to cite improved product and service innovation.

Governance gaps are creating trust problems with AI agents

One of the sharper findings concerns autonomous AI systems. Eighty-one percent of organizations expect AI agents to be making impactful business decisions within the next 12 months. Only 25% currently say they fully trust AI systems operating without human oversight. That is a meaningful gap for any operations leader who needs to define where AI can act independently and where it cannot.

On the governance side, only 33% of organizations have established clear policies defining the boundaries of AI decision-making authority. Twenty-seven percent are using registries and monitoring capabilities across all their AI systems. Sixty-one percent have redesigned at least some roles to account for AI, and 24% are standing up new AI-focused management functions.

Kyndryl's Chief Human Resources Officer Mark Paulek framed the underlying challenge in the report: organizations that are pulling ahead are aligning employee skills, role definitions, and decision-making authority to reflect how work is actually changing, not how it used to operate.

What this means for your team

• Audit your AI objectives against deployment scope. If your organization is in the 57% with broad AI deployment but not in the 32% hitting key objectives, the gap is most likely a people and governance issue, not a technology one.
• Benchmark your governance posture against the Pacesetter model. Policy coverage for AI decision boundaries, system registries, and monitoring are the specific gaps the Kyndryl data calls out. Check whether your frameworks address all three.
• Treat AI agent autonomy as an urgent governance question. With 81% of enterprises expecting agentic AI to make significant decisions within the year, defining human-oversight thresholds now is a compliance and operational risk issue, not a future planning exercise.
• Tie upskilling investment to role redesign, not just training headcount. The Kyndryl data shows that organizations achieving outcomes are restructuring how work gets done, not just adding courses to an LMS.