Fidelity Investments is the eighth named company in the May-2026 cohort to cut tech headcount and the first to do it while insisting, with a straight face, that AI had nothing to do with it.
On May 7, the family-controlled Boston firm — eighty-thousand employees, Abigail Johnson in the chair — announced it would eliminate about 1,000 positions, roughly 1% of the global workforce, and queue 3,300 new hires this year plus another 2,000 already-open positions still to fill. Net of attrition the company is growing the headcount, not shrinking it. The cuts concentrate in technology and product, the same place every other May layoff has concentrated. Effective June 1, the existing “agile squad” structure is being scrapped in favor of larger, more integrated teams; senior leadership layers are being trimmed.
The line every other company in the cohort would have said here is some variant of “AI is a new utility / AI-first / AI-native / over half the code is AI.” Fidelity, alone in the cohort, said the opposite: per Bloomberg’s confirmation, AI played no role in the decision.
The Boston Globe’s follow-up on May 11 ran with the headline “Fidelity is growing its work force by thousands. Blame AI.”
What’s actually being said
Read the two Globe pieces together and Fidelity’s position is internally consistent in a way the cohort’s other eight aren’t. The May 7 framing is “this is a tech operating-model change, not an AI restructuring.” The May 11 framing — Larry Edelman’s Trendlines column — is “exactly: AI couldn’t do this work yet, so we have to hire humans to do it.”
The 2,000-early-career-engineers hire is the part that gives the second framing teeth. Early-career software engineers are precisely the cohort that Anthropic CEO Dario Amodei warned could see up to half of all white-collar jobs eliminated within five years, and that Ford CEO Jim Farley issued the same forecast on in July. They are the cohort whose vulnerability is the empirical content of “AI-first.” If Fidelity were AI-first in the May-2026 cohort sense, the 2,000-person early-career hire would be the first slot to disappear.
Boston College economist Brian Bethune, quoted in the Globe, gives the cleanest version of the read: “The problem with AI is that there is a massive supply bottleneck [even though] they are slapping up data centers as fast as they can. Some companies may be shifting back to hiring smart, bushy-tailed [computer science] grads to move things forward instead of waiting in the AI queue.”
That is the entire thesis: Fidelity is hiring around an AI bottleneck, not in flight from an AI tailwind.
The May-2026 cohort, restated
After Fidelity, the cohort looks like this — ranked by stated framing:
- Coinbase, May 5: 700 cut, ~14%, “AI-native.”
- Ticketmaster, May 6: 350 cut, ~8%, “AI as a new utility.”
- Cloudflare, May 7: 1,100 cut, ~20%, “AI-first,” internal AI usage up 600% in three months.
- Microsoft, May 7: 8,750-eligible Voluntary Retirement Program, “Rule of 70.”
- Freshworks, May 7: 500 cut, ~11%, “more than half the code is AI.”
- BILL Holdings, May 7: up to 30%, “AI-native,” paired with a $1B buyback in the same release.
- PayPal, May 7: 20% over two-to-three years, $1.5B savings target.
- Upwork, May 7: ~24% (~145), Q1 flat, stock down 19%.
- Fidelity, May 7: ~1,000 cut, 1% — 5,300 new hires (half tech/product), 2,000 of them early-career — “AI played no role.”
Eight of nine companies in the cohort framed the cut as AI-driven. The ninth, the largest of them by employee count, framed the cut as the opposite. The cohort thesis (“AI is doing the work”) and the Fidelity counter-thesis (“AI cannot yet do the work, so we are hiring humans to do it”) are mutually exclusive at the level of the public narrative, but they are not mutually exclusive at the level of the business decision. Both can be — and almost certainly are — partially true at the same firm.
What Fidelity has done that the other eight haven’t is decline to put the AI label on a cost-line move that runs in parallel with a much larger hiring move. The simpler way to say it: Amazon, Block, Meta, Coinbase, Cloudflare, Freshworks, BILL, PayPal, and Upwork are using “AI” as the narrative grease on a headcount-flat-or-down decision. Fidelity is making a headcount-up decision and so doesn’t need the grease.
The labor-market backdrop is darker than Fidelity’s number suggests
Edelman attaches two Bureau of Labor Statistics figures to the Fidelity story that deserve repeating:
- Information-sector layoffs hit 66,000 in March 2026 — the highest reading since the pandemic.
- The information sector shed 92,000 jobs in the year through April — about 3% of the sector.
And a Quinnipiac March 2026 poll found that 71% of employed white-collar workers and 73% of blue-collar workers expect AI advances to reduce job openings.
Inside that backdrop, Fidelity’s 1%-cut, 5,300-hire move is not a counter-trend signal so much as a single-firm timing call. The bet underneath it is that AI capability per dollar of compute will lag for long enough that 2,000 new early-career engineers will produce more value over the next 24 months than waiting for the model layer to deliver the equivalent. Block CEO Jack Dorsey took the opposite bet in February when Block cut 40% of its workforce — “rather than dragging out layoffs as the company inevitably increased its AI use.” Both bets cannot win.
What to watch
- Whether the 2,000-person early-career class actually gets hired. Fidelity’s recruiting calendar usually closes in August. If the September new-hire count meaningfully undershoots 2,000, “AI played no role” gets re-read as a Q2 communications choice rather than a Q3–Q4 operating reality.
- Whether the agile-squad-to-larger-team flip is itself AI-shaped. “Bigger integrated teams” is the same org shape every May-2026 AI-first company has converged to. The structural change can be AI-driven even if the layoff justification isn’t.
- Whether the cohort’s framing settles. If a tenth firm in May reuses the Fidelity counter-line — “AI played no role, we’re hiring around it” — it becomes the second branch of the May-2026 corporate dialect, alongside “AI-native.” The half-life of corporate dialect being what it is, watch the next thirty days.
The dryly funny part
The Globe’s headline is the most quotable line in the entire May-2026 cohort: Fidelity is growing its work force by thousands. Blame AI.
Eight companies cut their workforce and blamed AI. The ninth grew its workforce — and the same vocabulary still fits. The word “AI” is doing identical narrative work in both directions in the same news cycle. By the time the Q3 earnings reports land, half of this cohort will be re-classifying its May press release as “transition planning” and the other half will be reclassifying it as “AI-driven productivity gain.” The 2,000 early-career engineers Fidelity is currently interviewing will, by then, be six weeks into their first standup — held, per the new operating model, with a much larger team and noticeably fewer senior managers in the Zoom grid.