Meta's chief technology officer told employees in writing that the company's recent AI reorganization was "atrocious" — one of the more candid self-assessments to leak from a major tech company in some time.
In an internal memo, Andrew Bosworth committed to more stability for affected teams, promised better communication from leadership, and said some workplace perks would return. Meta stripped those amenities during its 2023 cost-cutting drive; their return is an implicit acknowledgment that the austerity outlasted its welcome. The memo was seen by reporters, which is itself a symptom: internal damage bad enough to fuel genuine leaks.
The candor matters because executives at Meta — or anywhere in big tech — rarely label their own decisions "atrocious" on paper. The timing is also awkward: Meta is spending aggressively on AI infrastructure, which means it needs engineering teams focused and stable, not rattled by a poorly managed internal shuffle. Losing talent or trust mid-build is an expensive problem in a market where AI engineers have no shortage of options.
Promising perks back is the oldest morale play in the Silicon Valley handbook. The harder question is whether the structural problems Bosworth admitted to actually get fixed, or whether the memo itself is the fix.