July 3, 2026 • 6 Min Read

The 95% Problem Nobody Wants to Name

Artificial intelligence has never been easier to buy, build, or deploy. Realizing business value, however, remains the greatest challenge. The MIT NANDA initiative found that 95% of enterprise generative-AI pilots deliver no measurable P&L impact. Our own research reinforces this gap between ambition and results: only 14% of CFOs report clear, measurable impact from their AI investments today, yet 66% expect to achieve significant positive ROI within the next two years (RGP, 2026).

Leaders should ask what that failure rate is really telling them. AI is not falling short because technology doesn’t work. It falls short when people don’t embrace new ways of working.

The organizations creating the greatest value from AI are not those that moved first. They are the ones that prepared, engaged, and equipped their people to succeed.

AI Transformation Is a Behavior Change Wearing a Technology Costume

Most organizations approach AI investments like traditional IT projects. They implement the platform, integrate the data, train a handful of power users, and consider the work complete. But AI is not just a technology implementation. It changes how people think, make decisions, and work. That makes AI one of the most significant people transformations an organization will ever undertake.

When the people side of AI is neglected, business value suffers. Adoption stalls, employees revert to old behaviors, managers struggle to lead through the change, and informal workarounds become the norm. The result is an AI investment that delivers far less value than intended.

The evidence of what closes that gap is compelling. Gartner’s 2025 benchmarking research found that companies who have achieved successful change management report 2x higher revenue growth. Change management is not a layer added after AI is deployed. It is the foundation that determines whether the investment delivers lasting business value.

A Decade of Change in Three Years

AI has compressed what once took a decade of workplace change into just a few years. Technology has never been more capable, but the human side has never been more challenging. AI accelerates change faster than organizations can naturally adapt, leaving little time for people to adjust. The constraint is no longer technology or computing power. It is human judgment, trust, and the willingness to embrace new ways of working.

When organizations overlook that reality, the consequences are significant. AI initiatives stall before delivering business value. Employees disengage. Confidence in AI-generated outputs declines. Over time, a divide emerges between those who embrace AI and those who quietly return to familiar ways of working. The result is slower adoption, inconsistent outcomes, and a lower return on AI investments.

A Path Through: Align, Engage, Sustain

Closing the AI adoption gap is not a mystery. It requires a deliberate, structured approach. At RGP, we guide organizations through enterprise AI transformations using our proprietary framework, built around three phases: Align, Engage, and Sustain.

Align. Start with strategy, not software. Establish a clear AI vision, align leaders around the business outcomes that matter most, prioritize the use cases that create measurable value, and understand how work and roles will change. Visible executive sponsorship is one of the strongest predictors of adoption, yet it is often the first step organizations overlook.

Engage. Meet people where they are and help them move forward. Adoption is a journey from awareness to understanding and commitment, and stakeholders are spread across that journey at any given time. Understand where they are, then tailor communications, leadership engagement, and role-based enablement to move them forward. With change leaders activated across the business, organizations build the knowledge, confidence, and trust people need before AI becomes part of how they work.

Sustain. Adoption without reinforcement fades. Embed new behaviors through manager enablement, hands-on support during go-live, continuous feedback, and ongoing reinforcement long after launch. Reinforce progress with clear governance, defined ownership, and adoption metrics that measure behavior, not license counts, giving leaders the insight they need to sustain adoption and maximize business value.

What Separates the Winners

The payoff for getting this right is substantial. According to Gartner, organizations with successful AI initiatives spend up to 4 times more, as a share of revenue, on data quality, governance, and AI-ready infrastructure than organizations with underperforming AI efforts.

That connection is hard to ignore. High performers did not create value by deploying AI alone. They created value by transforming work, reshaping roles, and changing how decisions are made across the organization.

Where to Start

You do not need a fully developed AI strategy to get started. You need to design for adoption from the outset. These three questions will help determine whether your next AI initiative is positioned to deliver real business value.

First, have you defined success in human terms? If your only measures are technical, such as model accuracy, uptime, or licenses provisioned, you have no way of knowing whether people are working differently.

Second, do your people trust the tool, and do they understand what it means for them? Trust and role clarity are not secondary concerns. They determine whether adoption happens at all.

Third, have you planned to sustain the change rather than simply launch it? Reinforcement, governance, and clear ownership are what sustain adoption long after the rollout team has moved on.

Speed is easy. Adoption is hard. And today, adoption is where business value and competitive advantage are created.

RGP helps organizations turn AI investment into workforce adoption, and adoption into business value. To learn how our Change & Transformation Management practice can close your adoption gap, contact us today.

Sources:
MIT, NANDA Study, 2025
Gartner Report, 2025