Nav Thethi, well recognized and reputed Top DX Contributor, Executive Advisor, Podcast Host, Corporate Trainer, and Mentor.
AI is not failing because the models are weak. AI is failing because organizations are.
CX teams keep buying tools, launching pilots and layering chatbot strategies on top of legacy operational debt while ignoring the one discipline that actually determines whether AI scales or stalls: enterprise program governance.
In a recent podcast episode, Antonio Nieto-Rodriguez, former chairman of Project Management Institute (PMI), puts it bluntly, “Most companies are not short on technology or ideas. They’re short on execution discipline. AI will magnify that gap.”
Per MIT’s analysis of enterprise AI failure patterns from 2025, 95% of AI initiatives fail to achieve measurable business value, and the picture becomes painfully clear: AI doesn’t reward the most innovative companies. It rewards the most disciplined ones.
And the organizations that have discipline—portfolio discipline, prioritization discipline, delivery discipline—are the ones with modern, empowered, strategic PMOs.
Moreover, according to the CXPA whitepaper “Leading CX Into the AI Era,” the biggest blockers to AI-enabled CX have nothing to do with models or data. They are failures of leadership alignment, cultural readiness, cross-functional orchestration and the inability to move past pilot purgatory.
Where The Real CX-AI Maturity Gap Comes From
During recent years, I’ve been in touch with several executives from global organizations, and I can tell by using the five-stage digital maturity model, the same pattern repeats every time:
- Digital Awareness: Teams talk about AI but cannot operationalize it.
- Digital Experimentation: The organization launches many uncoordinated pilots—“innovation theater”.
- Digital Integration: Tools exist, but workflows remain siloed; CX and AI operate without governance.
- Digital Optimization: Data, CX, operations and AI start to converge—the first signs of real value.
- Digital Transformation: AI becomes embedded into enterprise orchestration, with program governance driving value realization.
Every AI-driven CX success story appears only in Stage 4 and Stage 5 organizations.
And what distinguishes these companies from the rest? Not their CX toolkit. Not their budget. Not their technology stack.
It is the strength of their program management discipline, the ability to:
- Prioritize the right problems
- Sequence AI use cases based on value
- Govern AI ethically and cross-functionally
- Manage change at scale
- Link CX and AI decisions to business outcomes
Weak PMO = weak AI outcomes. Strong PMO = scalable AI impact.
Perhaps this is the reason the labor market reflects something none of us were expecting. The U.S. Bureau of Labor Statistics projects +6% employment growth (2024-2034) for project management specialists, a signal that organizations are doubling down on execution discipline, not reducing it. AI eliminates administrative work, not the strategic governance required to turn innovation into measurable business value.
This is precisely why PMO and PM roles are on the rise: As AI accelerates complexity, companies need stronger orchestration, prioritization and value realization than ever before.
PMOs: The Missing Engine Behind AI-Driven CX
The CXPA whitepaper makes one point very clear: AI moves faster than traditional CX teams can. Without cross-functional orchestration, AI becomes chaotic. Yet most organizations treat AI as a technology initiative, not an enterprise transformation that requires program architecture.
Nieto-Rodriguez’s project economy perspective reinforces this: “AI demands fewer projects, better governance, and stronger value realization, not more activity.”
PMOs are the only organizational structure already built to handle:
- Cross-functional dependencies
- Risk, governance, accountability
- Multiyear transformation programs
- Sequencing high-impact versus low-impact work
- Ensuring value realization beyond go-live
AI without a PMO is like building a skyscraper with no blueprint. You may pour concrete, but nothing stands.
5 Takeaways: How To Reduce The CX-AI Maturity Gap
1. Move From Pilots To Portfolios
Most AI pilots fail because they are disconnected experiments. Shift to a portfolio approach that aligns every AI initiative to business outcomes, customer journeys and operational constraints.
2. Make The PMO The Owner Of AI Value Realization
CX and AI teams can define opportunities, but only a PMO can sustain governance, prioritize resources and measure value in a structured way across functions.
3. Advance Your Organization Through The Five Digital Maturity Stages
This is exactly where digital maturity becomes the real differentiator. Companies stuck in Stage 2 or Stage 3 will always struggle.
Invest in data foundations, workflow modernization and enterprise change to get to Stage 4 and Stage 5, the only stages where AI scales—responsibly, repeatably and with measurable value for both customers and employees.
4. Redefine CX Roles For The AI Era
CX teams must evolve from data collectors to experience builders, and from journey designers to AI-enabled orchestrators.
This requires new skills: prompting, workflow automation, cross-functional governance and ethical oversight.
5. Cut Projects, Not Corners
Following Nieto-Rodriguez’s principle, “The future belongs to organizations that do fewer projects, but deliver them exceptionally well.”
Eliminate the 60% of initiatives that add activity but not value. AI thrives in focus, not chaos.
My Perspective
AI isn’t failing because the models are wrong. It’s failing because the organization is not mature enough to sustain the transformation required to make AI matter.
CX excellence in the AI era isn’t about tools. It isn’t about chatbots. It isn’t about personalization engines.
It is about leadership, governance and disciplined program execution—the kind of execution only strong PMOs and digitally mature organizations can deliver.
This article was originally posted on Forbes.com