Companies Are Losing 55% of AI Potential Value Due to Skills Gap

A stark new report has quantified the "productivity paradox" facing the global economy: while nearly everyone is using artificial intelligence, almost no one is using it well enough to matter.

According to the newly released EY 2025 Work Reimagined Survey, organizations are currently missing out on approximately 55% of the potential productivity gains offered by AI. The study, which surveyed 15,000 employees and 1,500 employers across 29 countries, suggests that simply buying the software is not enough; a massive "human readiness" gap is turning potentially revolutionary tools into mere digital toys.

"Skimming the Surface"
The report reveals a deceptive statistic: while 88% of employees now use AI at work, only 5-7% use it for advanced, workflow-transforming tasks.

The Reality: Most workers are using sophisticated AI models merely to draft emails, summarize documents, or conduct basic searches—tasks that offer marginal efficiency gains but do not fundamentally change business outcomes.

The Potential: True productivity spikes come from "deep work" applications—coding, data analysis, and strategic forecasting—which the vast majority of the workforce has not been trained to execute.

"Organizations are suffering from a 'fragile talent foundation,'" the report notes. "They have deployed the technology but have failed to build the culture or skills required to leverage it."

The Training Trap
One of the most alarming findings for HR departments is the "retention paradox" discovered by the researchers.

High Reward: Employees who received extensive AI training (defined as over 81 hours annually) reported gaining an average of 14 hours of free time per week, nearly double the median.

High Risk: However, these highly skilled employees were also 55% more likely to quit their jobs. This creates a difficult bind for employers: investing in deep AI training creates super-productive workers who immediately become high-flight risks in a competitive talent market.

The Rise of "Shadow AI"
Because corporate training programs are lagging, employees are taking matters into their own hands. The survey found that between 23% and 58% of employees (depending on the sector) are bringing their own unapproved AI tools to work. This "Shadow AI" phenomenon not only creates security risks but indicates a massive, unmanaged demand for better digital tools that standard IT departments are failing to meet.