87,714 Layoffs Blamed on AI. Why NVIDIA's CEO Isn't Buying It.

Jensen Huang thinks your AI layoff memo is lazy, and the data on his desk backs him up. Companies have blamed artificial intelligence for 87,714 job cuts so far this year — the single most-cited reason — even as the CEO of the company selling everyone the chips says the automation doesn't yet justify the claim. The gap between what leaders say AI is doing and what it is actually doing shows up twice more this week: in the six hidden hours workers spend cleaning up after their own AI tools, and in the very real threats now following executives home from the office.

• 🏗️ Jensen Huang Calls the AI-Layoffs Excuse "Lazy": Oracle cut 21,000 jobs and called it AI. The company selling the chips says the math doesn't hold up.
• 🤖 AI Saves Your Team 11 Hours a Week — They Spend 6 Cleaning Up After It: A 6,000-worker study found the real AI tax hiding inside the productivity numbers.
• 🧠 Executive Security Spending Tripled Since the UnitedHealthcare Killing: Boards have quietly turned personal protection into a standard budget line.

Read time: 6 minutes 👇

Jensen Huang Calls the AI-Layoffs Excuse "Lazy"

Jensen Huang did something almost no one in Silicon Valley does: he called out his own customers. As CEOs lined up to blame AI for layoff after layoff, the NVIDIA chief executive said the logic was "lazy." Coming from the executive whose chips are powering the boom, the rebuttal lands differently than the usual skepticism.

Oracle's own regulatory filing admitted AI-driven restructuring cost the company 21,000 jobs over twelve months. Challenger, Gray & Christmas now tracks AI as the leading reason companies cite for job cuts in 2026, with 87,714 attributed to it year-to-date. But employers told NACE they expect to raise new-graduate hiring 5.6% this spring, and Salesforce is hiring 1,000 new grads and interns to build the very AI products it once used to justify cuts.

Why it matters: Whatever you say publicly about AI and headcount will be tested against reality within a year. The quiet-rehiring pattern that follows will cost you twice.

AI Saves Your Team 11 Hours a Week — They Spend 6 Cleaning Up After It

The number every vendor quotes is 11 hours saved per week. The number the Glean Work AI Index 2026 actually surfaced is 6.4 — the hours the average knowledge worker now spends feeding AI context, checking outputs, and rewriting confident-but-wrong answers. Glean's researchers gave the labor a name: botsitting.

Eighty-seven percent of workers use AI and say it saves them time, but only 13% of organizations perform measurably better for it. When botsitting goes unrecognized, 69% admit shipping AI work they cannot fully defend. At organizations that cited AI in layoffs, 94% of remaining workers ship unverified output, against 59% elsewhere.

Why it matters: If your AI ROI case rests on hours saved, you are measuring the wrong side of the ledger. The fix is redesigning who owns what in the workflow.

Executive Security Spending Tripled Since the UnitedHealthcare Killing

Two violent incidents in ten months rewired how boards think about the C-suite. Goldman Sachs Ayco's biennial survey found 27% of companies now provide personal security for their CEOs, up 59% in two years. Coinbase disclosed spending more than $6 million in 2025 protecting its chief executive.

Why it matters: This is no longer a discretionary perk you can defer to next year's budget cycle. If your board hasn't ordered an independent threat assessment for your senior leadership team, you are behind a norm that has shifted industry-wide in eighteen months.