Bill Pappas, Executive Vice President and Head of Global Technology and Operations, MetLife.
A company rolls out a powerful new set of AI tools. The pilot results are promising. Leaders invest in training, communications and change management. Then, reality sets in. Many employees simply don’t use the tools.
This pattern is showing up across industries. According to a July 2024 McKinsey survey, 78% of respondents reported that their companies used AI for at least one business function. However, a January 2026 Deloitte report noted that only 34% of surveyed enterprises had achieved meaningful, enterprise-wide transformation. The reality is the gap is about adoption, and it’s where the “last mile” of AI transformation either gets crossed or quietly stalls.
The last mile shows up when AI is technically deployed but not yet embedded into the way work actually gets done and baked into routines, decisions and team norms. Closing the gap is a leadership challenge that requires building confidence, setting expectations and reshaping habits so that AI becomes a trusted partner in daily work, not a tool that sits on the sidelines.
When Tools Arrive But Adoption Doesn’t
By the end of 2026, Gartner, Inc. predicts trillions of dollars will be invested in AI, including software, infrastructure, deployment and workforce upskilling. Yet in many organizations, the day after the rollout looks a lot like the day before. The tools are available, but the work hasn’t changed.
One reason is simple: People are opting out. A WalkMe study of 3,750 executives and employees (via Fortune) found that nearly eight in 10 workers (registration required) are ignoring or avoiding the AI tools their companies provide. For leaders, that’s the real bottleneck. Transformation only happens when employees trust AI enough to actually use it.
The Trust Barrier Behind The Behavior
When employees avoid AI, it’s easy to assume they need more training. Often, what they need first is reassurance. The decision to use AI has become personal. It’s shaped by questions of competence, autonomy and job security.
MetLife’s 24th Annual U.S. “Employee Benefit Trends Study” (EBTS) helps quantify what many leaders are hearing anecdotally. We found that 61% of responding employees are concerned about the ethical and safety risks associated with AI, including bias, misinformation and lack of accountability. Additionally, 59% fear AI will make jobs or skills obsolete faster than opportunities will emerge, and 24% feel like they need to compete with AI at work.
These concerns aren’t “resistance to change” in the abstract. They’re a rational response to a technology that can feel opaque and, at times, threatening. That’s why many well-intended AI programs stall: Organizations focus on tools, access, governance structures and enablement but miss a more foundational requirement. Employees need to believe AI is a force multiplier for their role, not a substitute for it. Trust is what ultimately connects deployment and day-to-day adoption.
A Leadership Playbook: The Five Rs
We’ve seen that closing the last mile requires leaders to be explicit about what AI means for people. The following five commitments (the “five Rs”) are a practical way to build confidence and move from pilots to scaled adoption.
1. Reassure employees that AI’s purpose is to augment their work, not augment them out of a job. This reassurance requires more than just an announcement at an all-hands meeting. It requires consistency, specificity and proof over time.
2. Reinforce the fact that AI isn’t for everything. It needs to be used purposefully, helping minimize the time it takes to complete mundane tasks so time and energy can be reallocated to higher-judgment work, relationship building and complex problem-solving. This will also help employees realize that leadership values their skills and isn’t replacing them.
3. Reward experimentation. Make a point to publicly and positively recognize the teams that try, iterate and share their learnings. Adoption accelerates when it’s modeled and celebrated, not just expected.
4. Regulate the use of AI with clear guardrails that address employees’ concerns. AI use needs human oversight, bias mitigation and ethical accountability, and employees need to see that this is being made a priority in practice, not just policy.
5. Respond and adapt. Be intentional about collecting feedback to understand employees’ needs and tailor support to address those needs. A workforce that feels heard is far more likely to engage.
Together, these five commitments can shift the conversation from deployment to genuine adoption and from compliance to confidence.
From Access To Adoption
For the past few years, conversations around AI adoption have centered around access to tools, assuming technology alone would drive transformation. However, even the most advanced systems will stay trapped in the pilot stage unless employees are willing to use them. Humans are hardwired to resist change, so implementing new technologies is often met with skepticism or pushback.
To respond to this resistance, leaders must actively reassure, reinforce, reward, regulate and respond—empowering employees and giving them the confidence they need to fully embrace AI and translate capability into real business impact.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
