The warning isn't coming from an academic — it's coming from the people advising Fortune 500s on how to use AI in the first place.
The Summary
- BCG's global head of people and org says half of C-suite execs surveyed already see critical skills eroding — and only 1 in 10 companies has a plan to address it
- The skills disappearing fastest: problem framing, judgment, creative thinking, solution evaluation — the exact ones leaders say matter most for long-term performance
- Two red flags of deskilling: using AI summaries as substitutes for reading, and inability to answer follow-up questions about material you supposedly reviewed
The Signal
David Martin runs people strategy for one of the world's top consulting firms. His job is helping companies figure out how to deploy AI without destroying their workforce's ability to think. When he says there's an "existential risk" of people deskilling themselves into unemployability, that's not an academic thought experiment. That's a market signal.
BCG surveyed 70+ C-suite leaders in June. Half reported active skill erosion. Not future concerns. Present tense deterioration. The kicker: only 10% of organizations have concrete plans to address it. A third haven't even discussed the problem. Companies are handing out hammers without teaching anyone not to smash their own thumbs.
"There is existential risk of people deskilling themselves to the point of being unviable in the market."
The pattern Martin sees: someone presents material they clearly didn't read. They used an AI summary. You can tell immediately because they fold under basic follow-up questions. They got the gist, missed the nuance, and now they're exposed in the meeting. This isn't about productivity tools making work easier. This is about cognitive outsourcing creating a gap between what you claim to know and what you can actually defend.
The irony cuts deep. The skills vanishing are the ones AI can't do well: framing problems correctly, exercising judgment in ambiguous situations, thinking creatively around constraints, evaluating whether a solution actually solves the real problem. We're offloading the exact capabilities that keep us valuable in an AI-saturated market.
Here's what's actually happening:
- Junior employees use AI to skip the learning curve, never develop baseline competencies
- Mid-level workers optimize for output velocity, lose depth in their domain
- Senior people delegate thinking to tools, atrophy their ability to spot flawed reasoning
Martin's first warning sign is the AI summary trap. Reading a summary isn't reading the material. You get conclusions without understanding how they were reached. You miss counterarguments. You can't connect dots the AI didn't connect. When someone challenges an assumption, you have no foundation to respond from because you never built one.
The Implication
If your employer isn't helping you navigate this, you're on your own. The market won't care that your company had no deskilling plan when your cognitive abilities have atrophied to the point where you can't compete.
The fix isn't abandoning AI. It's building guardrails. Use AI for first drafts, not final thoughts. Summarize complex material yourself first, then see how AI did it. If you can't explain something to a skeptical colleague without referring back to notes, you didn't learn it, you borrowed it. The agents are coming for everyone who mistakes borrowed intelligence for developed skill.