Memorandum
- From
- Caroline Mercer via Fast Company
- Date
- Filed
- Business·4 min to read
- Re
Companies invest billions in AI while employees struggle to deliver results
ReCompanies invest billions in AI while employees struggle to deliver results
A new survey finds a wide gap between worker confidence in AI tools and actual business outcomes, with many employees spending more time on AI tasks than doing them manually.
Companies continue to pour billions into artificial intelligence initiatives, yet a growing body of evidence suggests their workforces are not turning that investment into measurable business results. Gartner forecasts AI spending will reach $2.59 trillion this year, a 47% increase from 2025, but a study by Domino Data Lab found that the share of enterprises whose return on investment fails to outpace their spending has held steady at 57% since last year.
The disconnect appears to stem less from a lack of tools and more from a lack of support for the people using them. WalkMe's third annual AI at Work Pulse survey, which has tracked U.S. workers since 2024, found that while 90% of employees feel confident using AI, half said they have spent more time trying to get AI to complete a task than the task itself would have taken manually. The findings point to what researchers describe as an "AI confidence trap," in which workers overestimate their proficiency while underdelivering on efficiency gains.
Pressure from managers is compounding the problem. More than half of employees surveyed said their manager expects more output in the same amount of time, and 33% admitted to pretending they are more skilled at using AI than they actually are. The pressure extends beyond the office: Dice reports that AI skills now appear in 73% of tech job postings, and KPMG found that nearly half of companies are willing to pay an 11% to 15% salary premium for those skills. A PwC survey of 1,000 financial services executives found that 86% believe AI skills training is more important than an MBA for many new hires.
Leadership is not exempt from the trend. The WalkMe survey found that more than a quarter of managers have pretended to know how to use AI in a meeting or presentation, and 39% of senior leaders admitted to championing, approving, or purchasing an AI tool they do not know how to use. More than 40% of senior leaders said they have pretended to understand their company's AI strategy while communicating with employees, peers, or board members. The result is a top-down dynamic in which executives push adoption without grasping the tools themselves.
The burden of fixing the problem has fallen largely on IT and learning and development departments, but neither is equipped to handle it. Nearly a quarter of employees have received no AI training at all, and 39% report receiving conflicting messages about which tools they are allowed to use. Rather than enrolling in more courses, employees said they want guidance built directly into the tools, better integration with software they already use, and best practices tailored to their specific roles. When AI was effective, one-third of respondents said the reason was that guidance was available on their screen while working.
The survey suggests that organizations need to move beyond simply deploying AI and start measuring how it is used by individuals and teams. That means identifying where work breaks down, where efficiencies are discovered, and where AI actually slows performance. Building guidance and context into the tools themselves, rather than relying on separate training programs, appears to be the most effective path forward. As CIOs face increasing pressure to justify AI spending with hard data, linking employee usage to business outcomes has become a strategic necessity rather than an optional exercise.
