Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

When AI confidence can’t meet workforce reality

Three lessons for leaders.

[Photo: Getty Images]

Profile photo of Russ Fradin

CEO · Larridin

We’re living through one of the fastest workforce shifts in modern history. From customer service to engineering, nearly every knowledge worker is now expected or required to use AI. It’s embedded in daily workflows, built into software, and dominating boardroom conversations.

Companies are investing aggressively, but a common theme keeps emerging: identifying and measuring impact is more complicated than anticipated. Leaders are confident in AI, especially when they see the promise of improved speed and quality becoming real.

However, translating those perceived gains into a clear financial impact remains difficult. The result? Heavy investment, rising expectations, and a mismatch between leadership optimism and operational reality.

High-profile companies have experienced this firsthand and decided to reshape their strategies based on anticipated AI-driven efficiencies, only to reverse course when gains failed to appear. A 2025 Visier analysis of 2.4 million employees at 142 global companies found that these companies rehired 5.3% of their laid-off workers, and this rehiring trend is happening even with the rise in AI agent use.  

With this example in mind, companies can avoid learning their lessons the hard way. The difference between success and costly missteps lies in aligning executive confidence in AI with workforce reality, and designing a structure in which AI works as a multiplier.

Optimism vs. reality

The Larridin 2026 State of Enterprise Report underscores the disparity between leadership optimism and the reality of work. In the study, 92% of C-suite executives expressed full confidence in AI’s impact.

But how can that be true, when 58% of executives also reported unclear ownership of AI initiatives and only 38% maintain a complete inventory of AI tools in use? These findings reveal that executives are confident in AI’s value but lack visibility into how it’s actually being deployed, by whom, and how.

Research from Deloitte shows that only 11% of leading companies provide near-universal access to officially sanctioned AI tools. Among employees who do have access, fewer than 60% use AI tools in daily workflows.

This means that AI access doesn’t equal adoption, and executive optimism does not equal operational readiness.

Lessons learned

Workforce decisions aren’t about AI itself but about acting on projections rather than metrics.

Leaders naturally assume that technology-driven productivity gains will quickly offset workforce reductions. In reality, workflows and human-dependent strategic decisions are more complex. When executives see success with pilot programs, efficiency gains on specific tasks, and attractive, vendor-specific dashboards, the next step is to extrapolate organization-wide impact.

But Larridin’s study shows that adoption is uneven, governance is fragmented, and measurement is inconsistent. Nearly half of leaders admit that they don’t fully understand who is using AI, how much, or what for.

Making workforce decisions in this environment isn’t about AI failing to deliver value when properly used. It will. But are organizations ready to use AI effectively?

Most AI adoption challenges are organizational rather than technical. Without real-time visibility and insight on adoption and fluency, leaders can misjudge where AI is actually being adopted and how effectively it’s used.

3 AI essentials for leadership

Learning from the Salesforce scenario means gathering solid evidence before reshaping workforce strategy. Three leadership actions are essential.

1. Establish AI ownership

AI can’t live in organizational gray spaces. A designated executive leader must own AI adoption metrics, governance standards, and cross-functional alignment.

Many organizations appoint chief AI officers (CAIOs) to lead enterprise‑wide AI strategy and oversee the full portfolio of business-wide initiatives. The CAIO serves as the owner for AI governance, risk, and value realization, ensuring deployments are both responsible and measurably impactful. Accountability drives clarity.

2. Measure real adoption

Anecdotal success stories and surveys are insufficient. Leaders need system-level, real-time visibility into who is using AI, where it’s succeeding, and what outcomes it produces.

Workforce decisions should be guided by measured impact on adoption and efficiency (such as the number of active users, the complexity of prompts, time saved, etc.) and ROI—and organic attrition may be sufficient to make room for AI-powered productivity gains.

3. Integrate before you optimize

AI must be embedded into workflows through training, incentives, and process redesign. Only when AI usage becomes habitual and effective, meaning that employees have integrated these tools into their workflows and are experiencing productivity gains, should leaders consider structural workforce shifts.

AI matters and people do, too

This era of unprecedented AI momentum and investment offers a powerful opportunity to strengthen organizations, provided it’s guided by the right metrics, discipline, and instrumentation. AI is not a nice-to-have technology for presentations at board meetings. It’s a system demanding visibility, consistent execution, and accountability.

Organizations redesigning work processes with AI are twice as likely to exceed their revenue goals, per Gartner’s 2025 survey. The key question: By how much, and how quickly, can companies outperform through the strategic use of AI?

When leaders align ambition with operational reality, they lay the foundation for lasting advantage. To unlock human potential alongside AI, they must measure AI adoption, govern usage, and root workforce changes in their long-term strategic direction.

Russ Fradin is the CEO and cofounder of Larridin.