Meta CTO Calls AI Reorg 'Atrocious' in Staff Memo
Meta's chief technology officer called the company's AI reorganisation "atrocious" in a memo sent to staff on June 16 — and the word was not chosen by accident.
Andrew Bosworth, who joined the company then known as Facebook in 2006 and has been CTO since 2022, said Meta had done "an atrocious job explaining the vision" behind its new Applied AI division to the employees assigned to it.
What the Applied AI Reorg Actually Did to Meta's Workforce
The Applied AI unit was formed earlier in 2026 and swept up approximately 6,500 engineers and product managers, many of whom were given little choice about the transfer.
The move came alongside a May 2026 layoff of 8,000 employees — roughly 10% of Meta's 78,000-person workforce — and the reassignment of another 7,000 workers to AI training roles.
The layoffs landed just weeks after Meta reported $26.8 billion in net income for the first quarter of 2026.
The timing of the cuts, coming immediately after one of the richest quarters in the company's history, intensified the bitterness among those affected, IBTimes UK reported.
Several employees told Wired they had quietly hoped to be included in the layoffs, if only to secure the severance package on offer: 16 weeks of pay and 18 months of health cover.
📰 Related: Anthropic's Fable 5 Puts AI Reasoning Traces Under Scrutiny
'A Gulag' — What Workers Said About the Applied AI Division
The internal pressure on Bosworth to respond was real.
Workers in the Applied AI unit described the work as repetitive and creatively unsatisfying. One employee described the division to Wired as "a gulag."
TechCrunch reported on June 12 that the unit was "on the verge of revolt," with engineers reporting unclear roles, soul-crushing work, and record-low morale.
Bosworth had already acknowledged the atmosphere at an internal "Tuesdays with Boz" session on June 2, describing morale as "maybe not the worst it's ever been in 20 years here, but it's probably up there."
He compared the current atmosphere to the fallout from the Cambridge Analytica data scandal.
That comparison, from a 20-year company veteran, is not a small thing to put in writing.
📰 Related: KPMG AI Report Exposes a Bigger Problem Inside Corporate AI
🤖 AI Generated ImageWhat Bosworth Promised — and What It Actually Commits To
The June 16 memo, first obtained by Wired and cited across multiple outlets, paired the apology with specific structural commitments.
Managers will be capped at approximately 20 direct reports, a move designed to give employees more direct access to an advocate for their career growth.
Meta will also reduce how often workers are reassigned to new managers during reorganisations — which, for a division formed through involuntary transfers, is a direct acknowledgement of what went wrong.
Workers in the Applied AI unit will be allowed to apply for other roles within Meta rather than remaining locked in their transferred positions.
The memo also promised the return of workplace perks — restocked office microkitchens, increased travel budgets, and expanded spending on social events — that had been scaled back during the company's earlier "year of efficiency" cost-cutting phase.
The context behind all of this is financial scale. Meta has guided 2026 capital expenditure to between $125 billion and $145 billion — roughly double the $72.2 billion it spent in 2025. The cuts and transfers were framed as the workforce component of funding that investment.
📰 Related: Dario Amodei Calls for Government Power to Block Dangerous AI Models
The Talent War That Makes the Apology Strategic, Not Just Honest
Bosworth also used the memo to address a question employees had been raising directly: will AI replace their jobs?
His answer was careful. Meta does not believe AI will fully replace workers, he wrote, but added a caveat that employees should take seriously: "AI won't take your job, but someone who knows AI might."
The line lands differently inside a company that just laid off 8,000 people and forced 7,000 more into AI training roles.
What the memo reflects is the competitive reality Meta is operating in. Google, Microsoft, and Anthropic are all competing for the same pool of senior AI engineers. Meta knows it cannot win that competition with a demoralized workforce and an internal reputation — earned through a leaked memo in which a senior executive used the word "gulag" — that it is an unpleasant place to work.
Whether the manager cap, the optional reassignments, and the restocked snack kitchens are enough to reverse that reputation is the question the next 60 to 90 days will answer.
Key Takeaways
- Meta CTO Andrew Bosworth sent an internal memo on June 16 admitting the company did an "atrocious job" explaining its AI reorganisation to employees, first reported by Wired.
- The Applied AI division, formed in 2026, involuntarily transferred approximately 6,500 engineers and product managers; Meta also laid off 8,000 employees (10% of workforce) in May 2026.
- One employee described the division to Wired as "a gulag"; Bosworth himself compared morale to the post-Cambridge Analytica period at a June 2 internal session.
- The fixes include a manager cap of ~20 direct reports, reduced frequency of manager reassignments, and allowing Applied AI employees to apply for internal roles.
- Meta has guided 2026 capital expenditure to $125–$145 billion — roughly double 2025 spending — with layoffs and transfers framed as the workforce cost of that investment.
- Bosworth told employees "AI won't take your job, but someone who knows AI might" — a caution delivered inside a company that just removed 8,000 positions.
Sources
Also Read
Tech & AI Editor
David Park covers artificial intelligence, Big Tech, and the future of digital innovation. He translates complex tech developments into stories that matter for everyday readers.
