Ben Okonkwo: I have to start with a number that's been nagging me — 95%. Ninety-five percent of organizations have deployed AI in some form. That's the Gartner survey from July 27. And then you scroll two lines down and only 1 in 5 has gotten significant value out of it.
Marcus Vale: Hold on — that's the same survey where they say nearly a quarter of CHROs stopped junior hiring?
Ben Okonkwo: Same survey, Q4 2025, 110 HR heads. Published Sunday. And what that gap tells me is — the cutting was speculative. They cut because they feared AI would replace juniors, not because it demonstrably had.
Marcus Vale: Which is exactly why the reversal is so interesting — Lattice's Sarah Franklin restarted entry-level hiring. Booz Allen Hamilton is rebuilding headcount. Robert Half same story. The firms that moved fastest are already unwinding it. That's the episode.
Ben Okonkwo: Right — and the question I can't fully answer yet is whether Booz Allen counts cleanly here, because they also cut thousands of roles tied to federal contract losses under the Trump administration. So the AI-causation story is muddier for them than for Lattice.
Marcus Vale: Okay, but all three gave the same reason for reversing: AI cannot carry the full workload alone. That's not a budget story, that's a capability story.
Ben Okonkwo: Fair — so today we're really trying to figure out if what we're watching is a genuine market-wide correction or a few firms admitting they over-indexed.
Marcus Vale: And whether the 78% that never cut actually made the smarter call from the start.
Ben Okonkwo: But the 78% figure only makes sense if the 22% was real — and I'm not sure it is, at the scale being reported. The Gartner question was: does any business leader in your organization stopped hiring entry-level? That's one VP, one team. It's like saying one in four households has someone who stopped buying coffee — when actually it's one person in one household on a diet.
Marcus Vale: So the headline — 'nearly one-quarter of organizations' — is doing real work that the data doesn't actually support.
Ben Okonkwo: Right — and Human Resources Director Australia flagged exactly that on July 28. Their read was: in practice, few firms have actually stopped hiring for junior roles. That's the day after Gartner published. So you have a major finding and almost immediately a reality check from the field.
Marcus Vale: Huh. One day later.
Ben Okonkwo: One day. And the fieldwork was Q4 2025 — so the data is six to nine months old at publication. Now, I'm not saying that makes it wrong, but the gap matters if you're trying to track where hiring is right now.
Marcus Vale: Okay, but — if the 22% is soft, what do you do with the 95% adoption number? Because that one's from the same 110 HR heads. Same methodology problem, right?
Ben Okonkwo: Actually that's the one number I trust more, because 'have you implemented AI in some form' is a yes-or-no and you'd know the answer. The part that bites is the other half — only 1 in 5 says they've gotten significant or transformational value. So organizations installed AI almost universally, most of it isn't actually delivering, and some of them are cutting junior roles anyway. What are they replacing those juniors with, exactly? A tool that 80% of their own peers say isn't transformational yet?
Marcus Vale: That's the incoherence. The cutting logic assumes an ROI that the adoption data says hasn't arrived.
Ben Okonkwo: Kaelyn Lowmaster at Gartner put it plainly — redefine early-career roles, don't eliminate them, because cutting those pipelines risks workforce challenges three to seven years out that are really hard to unwind. That's the entry-level talent pipeline argument. And I think that's the right framing — not 'AI will replace juniors,' but 'what happens in 2029 when you have no mid-level bench because you didn't hire anyone in 2025.'
Marcus Vale: But that 2029 gap is exactly where the GMAC data gets uncomfortable. Because the GMAC report — cited in ETHRWorldSEA on July 28 — found 1 in 3 recruiters are already replacing junior tech roles with AI. Not thinking about it. Doing it. Singapore is in that cohort. So the pipeline isn't just theoretically at risk, it's actively being drawn down right now in specific markets.
Ben Okonkwo: That's the number that actually lands for me — because it's recruiters reporting behavior, not CHROs reporting what a VP maybe said.
Marcus Vale: Right. So the hot take isn't wrong — it's just wrong about geography. It's a tech-sector story, not an economy-wide one. And that distinction matters enormously.
Ben Okonkwo: Okay but now I want to pressure-test the Booz Allen piece of this, because — actually, this is the part that muddies the tech-specific reading. Their cuts came after the Trump administration slashed federal contracts. That's not an AI efficiency play, that's a revenue collapse. So when Booz Allen says AI can't carry the full workload and starts rehiring, how much of that reversal is about AI capability versus... just needing bodies again because the contracts came back?
Marcus Vale: That's a fair carve-out. Booz Allen is messier than Lattice here.
Ben Okonkwo: Lattice is the cleaner case — Sarah Franklin restarted hiring for a capability reason, not a contract reason. That's the signal.
Marcus Vale: Here's the concrete version of the pipeline argument — imagine a junior analyst hired in 2024. It's now 2026, she's eighteen months in, she's the person her team calls when the AI model surfaces something weird in the output. That's not a job the model does. That's a judgment layer you only have if you hired for it two years ago. Freeze hiring in 2024, that person doesn't exist. Kaelyn Lowmaster at Gartner is saying exactly that — you're not just losing a junior salary, you're losing the mid-level bench in 2028, 2029, 2031.
Ben Okonkwo: And three to seven years to close that gap is not a rounding error — that's a hiring cycle and a half.
Marcus Vale: Which is why the firms holding the line on junior hiring might actually be building something the cutters can't buy back easily — and that's the part we should get into next, because the augmentation framing changes the competitive math in ways that are harder to see right now.
Ben Okonkwo: And that competitive math is actually where the augmentation framing clarifies everything — because Gartner's own data supports augmentation as the dominant pattern, not replacement. That's not a reading, that's what Kaelyn Lowmaster said explicitly: redefine the early-career role so it contributes to higher-value work earlier. Not eliminate it. The AI handles the routine slice, the junior handles the judgment layer above it. That's a different job, not a missing job.
Marcus Vale: So the firms that cut — they misread the prescription.
Ben Okonkwo: They treated augmentation evidence as replacement permission. And those aren't the same thing. I mean — ninety-five percent adoption, one in five getting transformational value. That gap is what augmentation looks like at an early, messy stage. Not replacement.
Marcus Vale: The offshore parallel lands here, actually. Early 2000s — firms cut junior onshore roles for labor arbitrage, promised forty percent cost reductions, then spent years trying to rebuild the judgment they'd offshored away. Same mistake, different wrapper.
Ben Okonkwo: And the GMAC data — the one in three recruiters replacing junior tech roles — that's real, but it's sectoral. Tech-specific, Singapore-specific. Universalizing it is exactly how you make the wrong policy call.
Marcus Vale: Which means the 78% that held the line aren't laggards.
Ben Okonkwo: They may have — actually, no, I want to say this carefully — they may have been right by default, not by design. But the outcome is the same. They're building the bench the cutters will need to hire from in three to seven years. That's Lowmaster's pipeline argument made concrete.
Marcus Vale: So the defensible claim is: augmentation, not replacement, is what the data actually shows — and Lattice and Robert Half just proved it the hard way.
Ben Okonkwo: That's the one. Redefine the role. Don't eliminate it. The firms that figure that out now aren't behind — they're just not making a mistake that's already being quietly unwound elsewhere.
Marcus Vale: So the boldest AI move of 2025 was firing the 23-year-olds, and the boldest move of 2026 is hiring them back. Great pivot.
Ben Okonkwo: And the thing that actually gets me — Gartner polled 110 HR heads in Q4 2025, published July 27, 2026, and called it current. If the data's already six to nine months old and companies are already reversing mid-cycle... I mean, the real story isn't the 22%. It's how fast the correction is moving.
Marcus Vale: Yeah. We started with that 95% number like it meant something was decided. Turns out it just meant everyone bought the ticket.
