Why is AI making you busier instead of more productive in 2026?

Korn Ferry 2026: AI Adds Work for 52%

AI makes you busier when it multiplies tasks instead of judgment. Korn Ferry's Workforce 2026 report (September 8, 2026; 16,000 professionals, 11 markets) found 52% say AI increased their workload and 45% are too busy to deliver meaningful results. Naval Ravikant's rule is the fix: leverage only pays when it is applied to judgment, so audit every AI use against the decision it serves, not the output it produces.

The report that landed on my desk this week says, in numbers, what I keep hearing from the founders and executives I coach: the tools are faster, and the people are more tired. On September 8, 2026, Korn Ferry released its Workforce 2026 Global Insights Report, built on a survey of more than 16,000 professionals across 11 markets, from individual contributors to CEOs. Nearly two-thirds (62%) said their workload has grown significantly in the past two years, 45% said they are too busy to deliver meaningful results, and 44% said they are stretched beyond their capabilities. The line that matters for anyone using AI seriously is this one: 52% of workers said using AI has increased their workload, even though 63% said it improved their efficiency on the tasks where they used it.

Those two numbers are not a contradiction. They are the whole story. AI makes a task cheaper, so more tasks get assigned, attempted, or invented. The Korn Ferry data also shows the perception gap that produces this: 79% of CEOs reported AI-driven efficiency gains, versus 51% of individual contributors. The people setting expectations see the speed; the people doing the work absorb the volume. Meanwhile 61% now say they perform the responsibilities of more than one role, and employee motivation fell from 71% in 2024 to 61% in 2026. Korn Ferry's own summary is blunt: high activity does not translate into productivity.

I build AI systems and run them on myself every day, so I want to be honest about my own version of this. In early 2026 I let an agent draft every follow-up email, summarize every call, and pre-research every meeting. My output tripled. My decisions did not get better, and my evenings disappeared, because every artifact the agent produced still needed a human to read, correct, and act on it. I had multiplied the wrong thing.

The clearest lens I know for this comes from The Almanack of Naval Ravikant. Naval's formula is specific knowledge plus accountability plus leverage, and he is explicit that leverage without judgment is dangerous: "In an age of leverage, one correct decision can win everything," and a wrong one gets amplified just as fast. His line that knowledge workers should function like athletes, sprinting and then resting, is the opposite of what a 52%-busier workforce is doing. AI is the most permissionless leverage ever built, which is exactly why applying it to activity rather than judgment produces the Korn Ferry result: more motion, less meaning.

So I replaced my "automate everything" habit with a small weekly audit. I call it the Leverage Ledger, and it takes about twenty minutes on Friday afternoon. Step one: list every recurring AI workflow you ran this week, with the tool and the time it consumed, including review time. Step two: next to each, write the decision it serves, in one sentence; if you cannot name a decision, write "none." Step three: for each "none," cut it or cap it, because output nobody decides on is Korn Ferry's activity trap in miniature. Step four: for the workflows that survive, ask Naval's question, what is your aspirational hourly rate, and check whether the review time is worth it at that rate. Step five: reinvest the freed hours in one judgment task, a strategy memo, a hiring call, a pricing decision, and protect that block like an athlete protects recovery.

The first time I ran the ledger, four of my eleven AI workflows served no decision at all. Killing them freed roughly five hours a week, and nothing broke. The two workflows that survived with the highest score were both decision-shaped: a Claude project that red-teams my strategy drafts, and a Perplexity research pass before any hiring interview. Both produce fewer words than the ones I cut, and both changed what I did next.

Where this fails: the ledger does not help if your workload is set by someone else. If you are one of the 61% doing two jobs because a role was cut, the honest move is to take the ledger to your manager as evidence, not to optimize harder in private. It also does not fix the CEO-to-contributor perception gap by itself; that is a leadership problem, and the report's own suggestion, to shift from doing everything to prioritizing the work that matters, is only credible if leaders cut work rather than adding AI on top of it. And Naval's framing can tempt you to treat every task as beneath your hourly rate; some unglamorous work is where specific knowledge actually gets built.

What this changes about how you work is simple. Stop measuring AI by how much it produces and start measuring it by which decisions it improved. If the honest answer this week is "none," you have found the source of your busyness, and it was never the tool.

Sources: Korn Ferry, "Workforce 2026 Report: Unlocking Growth Requires Rethinking How Work Gets Done" (press release, September 8, 2026) — kornferry.com; HR Dive, "Korn Ferry: Leaders have to rethink how work gets done" (September 10, 2026) — hrdive.com; Allwork.Space, "New Survey Finds Global Workforce Working Harder, Producing Less" (September 9, 2026) — allwork.space; Eric Jorgenson, The Almanack of Naval Ravikant (2020).