On Culture: The Secret Life of AI at Work
- Myste Wylde
- Jun 11
- 6 min read
AI has developed a private life inside the company. Employees are using it, hiding it, refining it, judging it, depending on it, and sometimes resenting it before leaders have created the trust and rules to turn individual hacks into enterprise capability. That is where the culture risk gets interesting: the people finding the best AI uses may also be the least willing to make them visible.
One study found that 57% of employees admit hiding AI use, and daily users in low-trust organizations are nearly four times more likely to withhold AI knowledge than those in high-trust environments. Meanwhile, McKinsey reports that 88% of companies now use AI in at least one function, yet only 39% see material EBIT impact.
Workers are already saving real time, including 11 hours a week among digital workers. But some of that value is getting absorbed by botsitting, the new managerial burden of monitoring, correcting, and coordinating with AI tools instead of simply benefiting from them. Some disappears into tool switching, cognitive load, and unclear expectations. Frontline AI use has jumped to 74%, but the majority of employees receive limited or no guidance on what to do with the time saved.
This is where leadership matters. Trust determines whether employees share what they are learning or protect it as a private advantage. HR has to move beyond policy writing and help design the new architecture of human-agent work: where judgment belongs, where automation helps, where accountability lives, and how roles evolve. Managers need to translate saved time into better work, not just more work. Governance has to be clear enough to create confidence and flexible enough to keep pace with how quickly the tools are changing. The blunt question is: when employees discover a better way to work, do they believe sharing it will make them more valuable, or more vulnerable?
To the truth we earn,
Myste Wylde, COO
Why Employees Aren’t Transparent About Their AI Usage
Harvard Business Review By Eric Anicich and Jeslyn Brouwers
Summary: As AI moves deeper into daily work, a surprising culture risk is emerging: employees are hiding the very workflows that could create enterprise value. A global KPMG and University of Melbourne study found that 57% of employees admitted hiding AI use at work, and the authors’ own survey of daily AI users found that 30.3% had intentionally withheld AI-related knowledge, workflows, or techniques. The reasons are rational and revealing: employees fear being seen as less capable, assigned more work, giving up a competitive edge, or making themselves easier to replace. Trust changes the equation. Employees in the lowest quartile of organizational trust were nearly four times as likely to withhold AI knowledge as those in the highest quartile, 47% versus 14%, with a similar pattern for psychological safety, 45% versus 17%. The takeaway for leaders is uncomfortable but useful. AI ROI will depend less on tool access alone and more on whether people believe sharing what they have learned will strengthen their standing, improve the work, and earn credit rather than create risk. |
HR's Dual Mandate
McKinsey & Company By Sandra Durth, Bryan Hancock, Asmus Komm, and Ulf Schrader
Summary: Agentic AI is turning HR into one of the most strategic functions in the enterprise, with a mandate to redesign work across the organization while transforming its own operating model. McKinsey reports that 88% of companies now use AI in at least one business function, yet only 39% see material EBIT impact, largely because work, roles, workflows, governance, and capabilities have yet to be rebuilt around human-agent collaboration. The value at stake is significant: AI could create $150 billion to $200 billion in annual global value for HR alone, two-thirds of HR processes can be partially or fully automated, and current technologies could automate activities representing 57% of U.S. work hours, with 76% of jobs sitting in the messy middle that requires role redesign rather than simple replacement. AI transformation depends on HR moving from process administration to work architecture: shaping new role archetypes, reskilling at scale, clarifying decision rights, redesigning career pathways, and proving the model inside HR first so the function has the credibility to lead the enterprise through it. |
AI & the Productivity Paradox
Financial Times By Isabel Berwick
Summary: AI may be saving workers time before it creates measurable enterprise value. According to new Glean Work AI Institute research, digital workers report saving 11 hours a week with AI, yet only 13% see improved company performance. Much of the gap comes from the hidden labor around AI itself: workers spend 6.4 hours a week “botsitting,” or feeding tools context, checking outputs, rerunning prompts, and cleaning up mistakes; nearly 8 in 10 juggle multiple AI tools weekly; 60% run the same queries across different platforms; and one-third downplay the help they get from AI. The issue for CEOs is that time savings alone rarely translate into productivity unless data, workflows, team structures, incentives, and human oversight are redesigned around the work. AI layered on top of legacy processes may create more motion than momentum. |
AI at Work: Strategy Matters More Than Tools
Boston Consulting Group By Vinciane Beauchene, Sylvain Duranton, David Martin, Vanessa Lyon, and Jeff Walters
Summary: BCG’s fourth annual global AI at Work survey shows AI is changing jobs faster than companies are redesigning work to capture the value. Adoption has moved from the edge to the frontline: 74% of frontline employees now use AI every day or a few times a week, up 23 points from 2025, and 42% of regular frontline users save eight hours a week, the equivalent of a full workday. Yet 66% receive limited or no guidance on what to do with saved time, more than half fail to reinvest it into more strategic work, and only 28% see a strong connection between what leaders say about AI and what the organization actually does. The pressure is rising as well: 72% say skill expectations have shifted, only 36% feel adequately upskilled, 30% say AI agents are already integrated into workflows, and 61% believe agents could do at least half their job within three years. The advantage will come from strategic clarity, end-to-end workflow redesign, value measurement, governance, and visible leadership alignment so individual AI gains become enterprise performance instead of scattered time savings. |
AI is Creating a ‘Joy Paradox’ at Work
HR Dive By Lara Ewen
Summary: AI is making work more satisfying and more demanding at the same time: 67% of regular AI users say job satisfaction has increased, while 41% say cognitive load has also increased. The productivity gains are real, with 42% of regular frontline users saving a full workday each week, but the operating model has yet to catch up: 47% spend more time dealing with AI than doing actual work, 66% lack meaningful guidance on how to use the time saved, and more than half fail to reinvest that time into higher-value work. The shift is already material, with 72% saying AI has changed skill expectations, 30% reporting AI agents integrated into workflows, and 65% of managers and leaders believing agents will take over at least half their job within three years. Tthe opportunity sits in the messy middle between excitement and strain: clear strategy, stronger governance, better upskilling, and redesigned workflows will determine whether AI improves work or simply adds a new layer of cognitive load. |
Want the full newsletter each week in your inbox? Sign up now to save time and stay on top of trends.
Comments