PwC's Global Workforce Hopes and Fears Survey 2026, published September 29, 2026 from 49,364 workers in 48 countries, sorts people into front-runners (14%), AI insurgents (18%), indispensables (11%) and the engine room (56%). Your cohort rests on two habits: daily AI use (51% of front-runners versus 11% of the engine room) and scarce expertise. Daugherty and Wilson's Radically Human calls the pairing machine teaching: experts who teach the tools.

I read workforce surveys differently since I started coaching operators instead of managing them. The question is no longer "what should a company do about its people" but "which bucket would I land in if I answered honestly," and on this one the answer was less flattering than I expected.

The facts first, because the headlines compressed them. PwC published its Global Workforce Hopes and Fears Survey 2026 on Tuesday, September 29, 2026; it covers 49,364 workers across 48 countries and 29 sectors, surveyed in May and June 2026. Two scores sort everyone. An "AI advantage" score captures how much AI has raised the quality of your work, your creativity and skills, and your value to your employer. A demand score captures how scarce your skills are and how hard you would be to replace. Cross the two and you get front-runners (14%: high AI advantage, high demand), AI insurgents (18%: the heaviest weekly AI users, with less scarce skills), indispensables (11%: valued field expertise, little hands-on AI) and the engine room (56%: neither). Adoption overall is up, with 64% of workers using AI at work in the past 12 months, ten points higher than a year earlier, and daily generative-AI use rising from 14% to 22%. But the split inside that average is the story: 51% of front-runners use GenAI daily against 11% of the engine room, nearly 80% of front-runners have access to learning and development against under 40% of the engine room, and 75% of front-runners trust top management against 33%. PwC's global workforce leader Peter Brown told Business Insider the danger is that companies "render a big chunk of your workforce largely irrelevant in the world of work."

Two caveats that PwC's own write-up makes and most coverage dropped. The cohorts are partly self-reported: "AI advantage" is how much benefit you say you get, and "demand" is how in-demand you believe you are. And the survey does not establish causation. Daily AI users report more job security, but confident people in growing roles may simply pick up new tools faster. So I do not read the 14% as a class of gifted people. I read it as a description of two habits.

This is where Paul Daugherty and James Wilson's Radically Human earns its place on my shelf. Their central claim is that the next stage of human-machine work is not machine learning but machine teaching: domain experts imparting what they know to the systems, the way Shell's engineers teach its drilling models and Etsy's taste experts teach its style recommendations. The front-runner profile is exactly that pairing: scarce expertise, which is the demand axis, plus daily hands-on AI use, which is the advantage axis. The indispensables have expertise worth teaching and are not teaching it to any machine. The insurgents are teaching machines all day but have not yet built expertise worth the lesson. The engine room has access to neither, and PwC's data suggests that is largely a situation problem rather than an ability problem: only a quarter of engine-room workers learned new tools or skills last year, while two-thirds of front-runners say change arrived faster than ever and most of them applied new skills and took on more responsibility.

What I now do every quarter, for myself and with the founders I coach, is a four-step Front-Runner Audit built from PwC's two axes and Daugherty and Wilson's machine-teaching idea.

  1. Score both axes in writing, 1 to 5 each. Demand: if I disappeared for a month, how long until my judgment was replaced, and by whom? Advantage: in the last ten working days, how many of the hardest tasks did AI materially improve, not just speed up? PwC found that only 34% of daily users say AI helps with the most difficult parts of their job; be that honest.
  2. Name the one scarce thing and the one daily tool. Not a list. One piece of expertise clients pay for that a model cannot yet reproduce, and one AI tool you will open before email every day. If you cannot name the first, you are an insurgent; if you cannot name the second, an indispensable.
  3. Teach the machine for 20 minutes a day. This is the Radically Human move: write down your judgment calls as instructions, examples and rejected drafts inside a project or custom-instruction space, so the tool learns your standard rather than the internet's average. Expertise that stays in your head scores on the demand axis only.
  4. Ask for the reward out loud. Front-runners are nearly twice as likely as average to work somewhere that rewards effective AI use, and two-thirds say they will ask for a raise or promotion in the next 12 months. If your organisation rewards caution over experiment, that is data about the organisation, and 29% of front-runners say they are very likely to change employer within a year.

The limits, plainly. The audit is a mirror, not a ladder; it tells you which habit is missing, not how to acquire scarce expertise, which takes years. Front-running has costs PwC is candid about: this cohort experiences the fastest change and nearly half of them say they find it difficult to manage. The 62% AI-skills wage premium PwC cites comes from its separate AI Jobs Barometer and measures advertised roles, not raises for incumbents. And the daily users themselves are the most sceptical group in the survey: 31% name accuracy and quality as their top barrier, six points above average. Heavy use that produces discernment is the front-runner trait; heavy use that produces faith is just the insurgent trait with a better job title.

Sources: PwC, "Global Workforce Hopes and Fears Survey 2026," published September 29, 2026 (pwc.com/gx/en/1/issues/workforce/hopes-and-fears.html); Polly Thompson, "PwC's huge survey of 50,000 workers says AI is splitting employees into 4 distinct groups," Business Insider, September 29, 2026 (businessinsider.com); The Next Web via daily.dev, "PwC survey of 49,000 workers finds most are falling behind on AI skills," September 29, 2026, cohort percentages (daily.dev); Paul R. Daugherty and H. James Wilson, Radically Human (Harvard Business Review Press, 2022), chapters on Expertise and Talent.

Related: How to Find Your Passion · Best Self-Improvement Books · How to Make Better Decisions · What University Will Not Teach You