What: Seramount's analysis of more than 10,000 US employee responses finds that workplace culture, not AI adoption rates, determines whether AI produces business value.

Why it is important: Usage and time-saved metrics do not show whether AI is improving results. Value depends on redesigning work, preparing managers and deciding how saved time is reinvested before tools scale.

Seramount's report draws on more than 10,000 employee responses collected between 2021 and 2025 and weighted to 2024 U.S. Bureau of Labor Statistics benchmarks, alongside interviews with eight HR leaders. Citing McKinsey and Deloitte data, it shows that AI use is widespread but rarely converts into business value: 88% of organisations report regular AI use, yet only 30% are redesigning key processes around it.

The workforce data points to weak foundations. Only 32% of employees understand what it takes to succeed in their role, and 41% feel comfortable taking risks without fear. The report warns of "quiet compliance", where employees use the tools without questioning outputs or measuring whether work is improving. Saved time is absorbed rather than reinvested: Glean data shows AI saves workers 11 hours a week, but only 13% say their organisation performs significantly better as a result.

Seramount sets out five conditions for AI value: clarify how roles and performance measures will change, redesign workflows and decide how saved time is used, connect AI fluency to learning and sponsorship, equip managers to translate strategy into daily expectations, and make it safe to challenge AI outputs. It recommends tracking work quality, workload and psychological safety alongside adoption.

IADS Notes: The gap between AI adoption and measurable results has been building across retail and workforce research. Faster output was already found to coexist with heavier review burdens, weaker feedback loops and an erosion of the judgment that sustains performance when work is not redesigned (Seramount, March 2026). The same month, 50% to 55% of US jobs were projected to change materially within two to three years, with only 10% to 15% at risk of elimination, which puts role redesign and upskilling ahead of headcount reduction (BCG, April 2026). Walmart has since rolled out role-specific AI certifications to its 2.1 million associates and framed the technology as augmenting jobs rather than replacing them (CIO Dive, April 2026). On culture, executives were shown to overestimate employee enthusiasm for AI, and psychological safety was identified as the condition that lets staff question outputs and admit confusion (Seramount, June 2026). Middle managers were then found to be where AI initiatives most often stall, with five distinct manager profiles each requiring a different intervention (Harvard Business Review, September 2026).

U.S. workforce data shows culture, not adoption, decides AI's business value