Speed is not the same as judgment, and companies are starting to notice the difference
Walk into almost any boardroom discussion about artificial intelligence and you will hear a familiar assumption. The young people will lead this, because they grew up with screens, prompts and apps. The older colleagues will need to be brought along, gently, and preferably before they retire.
It is an understandable belief, and it is increasingly looking wrong. When AI leaves the demo stage and enters real work, with real customers, real liability and real consequences, the question is no longer who uses the tool most often. It is who knows when the tool is wrong. On that question, experienced professionals and AI turn out to be a far stronger combination than most organisations have planned for.
The usage gap is real, but it measures the wrong thing
Younger people do use AI more often. In Germany, Bitkom’s 2026 survey found that 29 percent of 16- to 29-year-olds use AI every day, compared with 22 percent of 30- to 49-year-olds. In the United States, the AARP Employer Resource Center reports that workers over 50 use generative AI at work at roughly half the rate of colleagues under 40, 17 percent against 34 percent.
But frequency tells you very little about quality. Someone who asks a chatbot to rewrite twenty emails a day is a heavy user. Someone who asks it three precise questions about a contract risk, checks the answers against what they know, and spots the one claim that does not hold up is a sophisticated user. Organisations need far more of the second kind.
And the same AARP analysis is clear about why older workers lag on usage: more than half received no technology training from their employer in the past year, compared with 27 percent of younger workers. That is not a talent problem. It is an investment decision, and arguably a poor one.
What the research says about seniority and AI sophistication
One of the most revealing studies so far comes from inside a large professional services firm. Researchers Nicholas Hallman, Zachary Kowaleski, Anu Puvvada and Jaime Schmidt analysed more than 713,000 real prompts from nearly 4,000 KPMG employees during 2025. They measured how clearly people wrote their prompts, whether they used deliberate techniques such as asking the AI to check its own work, and how broadly they applied it.
On every measure, sophistication rose with seniority. Staff scored lowest, managers higher, and those above manager level highest. Senior employees also used AI for more knowledge-intensive work, while junior staff leaned towards writing tasks. The authors suggest that seniors may simply be better at delegation: setting objectives, reviewing intermediate work and pushing back until the output is right. In other words, the skills of a good manager are the skills of a good AI user.
The same paper notes that wider survey evidence shows AI use at work peaking among people in their 30s and 40s rather than among the youngest workers, which further undermines the “digital natives will lead” narrative.
Why context is the real competitive advantage
Generative AI is very good at producing fluent, confident answers. It is much less good at telling you when its answer is plausible but wrong. That gap is exactly where experience earns its keep.
The American Psychological Association put it simply in a 2026 review of AI and human skills: experts are generally better equipped to critically evaluate AI outputs, while novices may struggle. An experienced lawyer will notice flawed reasoning that a first-year law student will accept.
The landmark Harvard and Boston Consulting Group experiment with 758 consultants shows how costly that difference can be. On tasks well within AI’s capabilities, consultants with AI worked faster and produced better results, and the weakest performers gained the most. But on a task deliberately designed to sit just outside what the AI could do well, consultants using AI were 19 percentage points less likely to reach the correct answer than those working without it. The tool did not fail loudly. It failed persuasively.
Follow-up research from the same team found that when consultants challenged the AI, it often doubled down with more data and more confident language, a pattern the authors call “persuasion bombing.” Resisting that takes something no training module can deliver quickly: a deep, lived sense of how things actually work in a market, a client relationship or a production line.
German labour market researchers are seeing a similar pattern from a different angle. The Institute for Employment Research (IAB) found that between 2022 and 2025, employment in AI-exposed occupations grew faster than elsewhere for workers aged 30 to 64, while it developed more weakly for those aged 20 to 29. As DIE ZEIT reported, the IAB explains this partly by older workers’ greater experience-based knowledge and stronger professional judgment.
Case study: Ford and the return of the “grey beards”
Few companies illustrate the point more vividly than Ford. In 2023, the carmaker leaned heavily into AI-driven quality control, relying on more than 900 AI-assisted cameras to identify defects at scale. The expectation was that technology and revised design requirements would keep quality high.
They did not, at least not on their own. Defects persisted, and Ford concluded it had over-corrected. The company brought back, hired or promoted roughly 350 experienced technical specialists, veteran engineers known internally as “grey beards.” Their job was not to replace the AI but to train it better, lead design reviews, hunt for failure points before parts reached the plant floor, and mentor younger colleagues.
Charles Poon, Ford’s vice president of vehicle hardware engineering, was candid about the lesson, telling Bloomberg that “mistakenly we thought that by just introducing artificial intelligence and ingesting the design requirements that we had, that that would produce a high-quality product.” He added that the company needed its AI tools to be “trained by the most experienced individuals” (Brian Solis).
The results were tangible. Ford ranked highest among mass-market brands in the 2026 J.D. Power U.S. Initial Quality Study, its first number-one finish in that category since 2010, and CEO Jim Farley credited the move with saving “hundreds and hundreds of millions of dollars” in warranty and recall costs. The AI was always capable. It needed people who knew what good looked like.
Tacit knowledge: the part of the business that was never written down
There is a reason experienced people catch what AI misses. Much of what they know has never been documented. It lives in pattern recognition, in memories of the deal that went wrong in 2009, in a feel for which supplier will quietly miss a deadline.
Stefan Meffert, managing director of the German metal housing manufacturer apra-norm, described this well when his company joined a research project to capture expert knowledge with AI. The workflow changes every day, he told aktiv, and “this knowledge cannot simply be written down.” The AI could only learn because experienced employees were there to teach it.
For German companies, this is about to become urgent. According to the Institut der deutschen Wirtschaft, the country’s potential workforce will shrink by around 4.3 million people by 2036 as baby boomers retire. A separate IW/KOFA study counts 8.5 million employees aged 55 or older, a quarter of the workforce, who are expected to leave within the next ten to fifteen years. If their judgment is not built into how AI is deployed now, it will simply walk out of the door.
The confidence gap that companies are getting wrong
Bitkom’s latest company survey shows that 57 percent of German firms now use AI, up from 36 percent a year earlier, yet two thirds rate their employees’ AI skills as low. The temptation is to fix this by putting the youngest and most enthusiastic people in charge.
That would be a mistake. As communications expert Bärbel Hestert-Vecoli argued in Handelsblatt, companies lean too heavily on younger staff during transformation and overlook the experience of the baby boomer generation, despite the widespread but mistaken belief that people over 55 are less digitally capable. The data suggests they are catching up fast: a joint AARP and LinkedIn study found that the number of workers over 50 listing AI and related technologies as skills grew by 25 percent over five years, nearly double the rate for younger workers.
This is not a generational contest
None of this means younger professionals are poor AI users, or that seniority guarantees good judgment. The BCG research is clear that AI lifts less experienced workers substantially on routine tasks, and every organisation needs people who learn new tools quickly and experiment without fear.
The point is about roles. Young talent often brings speed and fluency. Experienced professionals bring context, scepticism and accountability. The most effective model pairs the two deliberately: younger colleagues showing what the tools can do, and senior colleagues deciding what they should do and checking whether they have done it. Harvard researchers even titled one of their papers “Don’t Expect Juniors to Teach Senior Professionals to Use Generative AI”, after finding that well-intentioned junior advice often misjudged where the real risks lay.
What this means for leaders and boards
For executives shaping their organisation’s AI strategy, a few practical conclusions follow:
Put experience at the review stage. Let AI draft, analyse and summarise, but make sure decisions with financial, legal or reputational weight pass through someone who has seen similar situations before.
Train your senior people first, not last. The usage gap among older employees largely reflects missing investment, not missing ability.
Capture tacit knowledge before it retires. Use experienced staff to train, test and correct internal AI systems while they are still in the building.
Build mixed teams on purpose. Pair digital fluency with domain judgment rather than hoping the two will meet by accident.
Measure quality, not just adoption. Logins and prompt counts say little. Error rates, rework and decision quality say much more.
For senior leaders themselves, the message is equally clear. Experience is not a liability in the age of AI. It is the asset that makes AI safe and valuable. WFA has explored how this is already creating new roles for experienced executives, from AI governance to board-level advisory positions, and why decades of judgment are something you turn into lasting contribution rather than quietly set aside. As restructuring reshapes management layers across Germany, the leaders who can tell a convincing story about how they combine experience with new tools are the ones who stand out, a theme WFA examined in The Title May Not Travel. The Leadership Story Can.
The bottom line
The organisations that get the most from artificial intelligence will not be the ones with the youngest workforce. They will be the ones that understand that experienced professionals and AI belong together, because AI produces answers and experience knows which answers to trust. The technology is getting smarter every month. Judgment, context and the courage to say “that doesn’t look right” are still built the old-fashioned way, over years.
Considering how your experience fits into an AI-shaped next chapter? WiseForce Advisors supports senior leaders through executive transitions with peer advisors who have held the roles themselves. Explore more perspectives in WFA Insights or get in touch for a confidential conversation.
