Deconstructing the "Pseudo-Prosperity" of Executive AI Adoption: Less Than 20 Minutes a Day with Limited Efficiency Gains

While the media remains obsessed with the grand narrative of "AI replacing humans," a sober empirical report has pierced the visionary bubble.

In February 2026, the National Bureau of Economic Research (NBER) released a significant working paper titled Firm Data on AI. The research team conducted an in-depth survey of 6,000 executives across the US, UK, Germany, and Australia, addressing the market’s most critical questions: How are companies actually using AI? Are executives using it themselves? Has it truly impacted performance?

The report reveals four clear business facts that every decision-maker should evaluate calmly.

1. Adoption Approaches 70%, Driven Predominantly by Generative AI

Across the four developed nations surveyed, AI adoption has reached 69%, with the US leading at 78%, followed closely by the UK and Germany. However, these high figures mask a highly imbalanced application logic: this is almost entirely a transformation driven by Generative AI alone.

The top three use cases focus exclusively on desk work: text generation (41%), visual content creation (30%), and data processing (28%). In contrast, autonomous driving (3%) and robotics (9%) remain marginal. This defines a clear reality: the current AI wave is primarily restructuring "knowledge and information" rather than "physics and manufacturing."

2. Executives Use AI, But Depth of Adoption Remains Low

The depth of a corporate transformation often depends on the firsthand experience of its leaders. Although the survey shows that over two-thirds of CEOs and CFOs interact with AI during a typical work week, this engagement remains highly "fragmented."

The data illustrates a stark contrast: executives average only 1.5 hours of AI use per week—less than 20 minutes a day. Approximately 41% of respondents use it for less than an hour weekly, while only 7% qualify as "heavy users." Although usage time has increased by 50% since early 2025, AI currently functions as a peripheral tool for specific tasks rather than a core component of strategic decision-making workflows.

3. Doubling Efficiency Through AI May Not Be Fact

In the eyes of the surveyed executives, AI has provided almost no substantive boost to corporate performance over the past three years.

The data directly debunks the short-term "efficiency myth": over 90% of firms stated that AI has not changed their hiring scale, and 89% of executives saw no change in productivity (sales per employee). Quantitative results show that AI's actual contribution to productivity growth across these four nations averaged a mere 0.29% over the past three years, with a near-zero net impact on total employment. Faced with macro-empirical evidence, the social media buzz surrounding "mass layoffs" and "doubled productivity" lacks factual support for now.

4. Rather Than a "Layoff Shock," AI is More Likely to Reduce Future Roles

Despite a quiet past, executive expectations for the "next three years" (through 2028) have shifted dramatically, signaling that a true tipping point is approaching.

Executives anticipate that AI will lift overall corporate productivity by 1.4% over the next three years (with US expectations reaching 2.3%). Accompanying this efficiency gain is a substantive streamlining of roles—a projected 0.7% reduction in full-time staff. Based on the current employment base of these four countries, this translates to the disappearance of roughly 1.75 million jobs.

Critically, about two-thirds of this reduction is expected to occur through "reduced hiring of new staff" rather than direct terminations. This suggests the impact will be extremely subtle: the pain lies not in being fired today, but in the job market’s doors quietly narrowing for tomorrow.

Conclusion

While current AI adoption rates are high, most applications remain surface-level "patchwork" at the tool level. The core logic of the next three years will shift from "experimenting with tools" to "restructuring processes." As productivity expectations begin to jump above 2%, enterprises must consider how to find a new balance between high efficiency and lean organizational structures.