As artificial intelligence becomes a board-level priority for organizations worldwide, a troubling disconnect has emerged between AI ambition and real-world results. While 77% of business leaders report AI as a strategic imperative, a staggering 94% face significant challenges deploying it at scale, according to new global research. Pete McEvoy, Global Head of DXC Technology‘s AdvisoryX Group, identifies this phenomenon as the “AI execution gap”�where organizations move faster on ambition than on the fundamentals needed to turn pilots into production-ready solutions.
In this exclusive VMblog Q&A, McEvoy reveals surprising findings from DXC’s latest research, including an unexpected lag in Japan’s AI adoption and a critical gap between where leaders think AI should be led versus where it’s delivering the most value. He explains how AdvisoryX’s integrated approach�combining strategy, operations, technology, and people�is designed to help enterprises move beyond isolated AI experiments to scaled, measurable impact. McEvoy also shares his outlook for 2026, predicting significant shifts in R&D, compliance, and front-office functions as organizations prepare for the next wave of AI-driven business transformation.
++
VMblog: You’ve talked about a growing “AI execution gap” in the enterprise. What’s driving this gap, and why do you think so many organizations struggle to turn AI strategy into real outcomes?
Pete McEvoy: The biggest driver of the AI execution gap is that organizations are moving faster on ambition than on fundamentals. AI has become a board-level priority, but many enterprises run into execution challenges spanning strategy, deployment focus, leadership alignment, organizational readiness, and technical capability to move beyond pilots.
Our research shows while 77% of leaders say AI is a board-level priority, 94% still face significant challenges deploying it at scale. What we see repeatedly is AI being launched as isolated initiatives, often driven by pressure to “do something with AI,” rather than tied to clear outcomes or embedded into how work actually gets done. Without optimized data, aligned leadership, and disciplined validation, pilots struggle to scale. Execution breaks down not because the technology doesn’t work, but because the enterprise isn’t ready to support it.
VMblog: Your team’s new global research points to some interesting contradictions in how leaders approach AI. What surprised you most about the findings?
McEvoy: The study found Japan’s AI adoption rate lagged behind expectations for such a highly industrialized nation. This, and other nuanced findings, point to the need for deeper analysis of how AI adoption varies across geographies, industries, and organizational capabilities.
The study also revealed a significant gap between priority and preparedness. While a majority of leaders say AI is a strategic priority, nearly two-thirds of organizations still can’t articulate a clear enterprise business case for it. That tells us adoption is often driven by urgency rather than outcomes. At the same time, nearly a third of organizations plan to implement agentic AI within months, yet many lack the foundational readiness to support it.
We also saw a discrepancy between where leaders believe AI should be led and where it is actually delivering the most value. Leaders often think AI should be led primarily by technical teams, yet adoption over the next three years is expected to accelerate fastest in business-critical areas like research & development, compliance, and environmental, social and governance reporting. These are functions where context, regulation, and decision-making are required, proving AI success depends as much on business leadership and organizational readiness as it does on the technology itself.
VMblog: You now lead AdvisoryX. What do you think distinguishes it?
McEvoy: What differentiates AdvisoryX is we’re built by design to close the execution gap we’re seeing across enterprises. We don’t separate strategy from execution or consulting from engineering-we bring them together from day one.
AdvisoryX combines DXC’s advisory capabilities across strategy, operations, people, risk, and user experience with our deep engineering expertise in areas like AI orchestration throughout the software development process. That allows us to help organizations diagnose challenges, design future operating models, and execute transformation at scale. AdvisoryX is supported by a research engine that moves at the pace of business, so our guidance reflects real-world conditions, not theoretical frameworks.
The goal is simple: help organizations build AI from the ground up in a way that scales responsibly and delivers measurable impact.
VMblog: AdvisoryX brings together strategy, operations, technology, and people in one model. How does this approach help organizations tackle the execution challenges you’re seeing?
McEvoy: Most AI initiatives fail because they’re treated as technology implementations rather than operating model changes. AdvisoryX takes a different approach. We focus on how AI fits into workflows, decision rights, and leadership models so AI can operate effectively alongside people.
As AI becomes embedded in business-critical functions, organizations need hybrid models where AI operates with partial autonomy and humans provide oversight and judgment. This requires redesigned processes, new skills, clear governance, and novel operating models. By integrating strategy, operations, technology, and people, AdvisoryX helps organizations move from experimentation to AI that actually works in practice.
VMblog: DXC also introduced five integrated AI solutions as part of AdvisoryX. How do these pieces fit together to help enterprises move from early experimentation to scaled impact?
The five solutions are designed to work as an integrated system, rather than standalone components, and address the most common points where initiatives break down:
- AI Core establishes the enterprise foundation with data, architecture, and governance.
- AI Reinvent brings proven use cases across human-assisted, semi-autonomous, and autonomous models.
- AI Interact focuses on redesigning workflows so people and AI can collaborate effectively.
- AI Validate ensures continuous testing, observability, and responsible governance.
- AI Manage supports production operations as models evolve.
Together, they are designed to meet organizations where they are, allowing our customers to start small, validate quickly, and scale repeatable AI solutions with confidence.
VMblog: In 2026, where do you see AI delivering the most impact for enterprises, and what should leaders be doing now to prepare?
McEvoy: We expect the most immediate shifts in how work is done to occur in functions where AI adoption is accelerating fastest and workforce demand is rising such as R&D, compliance, ESG reporting, and in technical areas like IT, data, cybersecurity, and software development. These are areas where leaders anticipate AI becoming more embedded in daily workflows and decision-making, not just in task automation.
Moreover, by mid to late 2026, AI adoption will shift to the front office organizations. New interfaces, interaction models, and commercial approaches will drive significant business transformation for early adopters.To prepare, leaders need to focus on execution readiness. That means aligning AI investment to measurable value, strengthening data and governance foundations, redesigning workflows, and preparing the workforce for new ways of collaborating with AI solutions.
##
