Mark Zuckerberg acknowledged that Meta platforms made substantive errors in the course of an aggressive corporate overhaul designed to redirect the company’s resources toward artificial intelligence , a concession that comes after one of the most disruptive job cut in the social media giant’s two-decade history.
In an internal communication reviewed by Reuters, the Meta Chief Executive Officer told employees the restructuring had not been executed without fault, and warned that further missteps remained possible as the company continued its AI-centric transformation.
“Given the complexity of these changes, we’ve made mistakes and will almost certainly make more,” Zuckerberg wrote, according to the memo. He added that his priority going forward was to provide organisational stability, while cautioning that external forces beyond the company’s control meant he could not make unconditional guarantees about the path ahead, reports cybernews.com.
The restructuring, announced in May 2026, involved the elimination of approximately ten per cent of Meta’s global headcount, a figure translating to around 8,000 positions. Concurrently, approximately 7,000 staff members were redeployed to newly created AI-oriented roles, a move that underscores the scale of the company’s pivot toward machine learning infrastructure and applied AI development.
The dual nature of the overhaul, simultaneous reductions and reassignments, created significant internal disruption, which Zuckerberg’s memo appears to address directly. He indicated that the company had deliberately structured the transitions to allow for some degree of reversibility, noting that misassigned employees could potentially be moved back to previous functions if errors in placement were identified.
Earlier reporting had indicated that Meta was considering a further round of redundancies in the second-half of the year. Zuckerberg has now pushed back against those expectations, stating that no additional company-wide job cuts are anticipated before the end of 2026. The clarification is likely to bring some relief to remaining staff, though the qualifier ‘company-wide’ leaves open the possibility of more targeted, departmental reductions.
The chief executive also flagged plans to invest more heavily in team cohesion, including an expansion of corporate off-sites and staff events. A large-scale hackathon is scheduled for July, intended to foster collaboration across newly formed AI units.
A separate concern raised in the memo relates to management ratios within Meta’s AI divisions. Reports had previously indicated that some of the company’s AI engineering teams operated with unusually flat hierarchies, with as many as fifty employees for every single manager in the Applied AI Engineering unit. Zuckerberg acknowledged that this had placed disproportionate administrative burden on those in supervisory roles and indicated the company intended to scale back such arrangements.
The disclosure points to a broader structural challenge: as Meta has prioritised technical headcount in AI, the organisational scaffolding required to manage such teams has lagged behind, creating bottlenecks and accountability gaps that may have contributed to the very mistakes the chief executive has now admitted.
The internal reckoning comes at a financially sensitive moment. Earlier this month, Meta’s share price declined more than five per cent following reporting by the Financial Times that the company was considering a multi-billion-dollar equity offering to help fund its AI ambitions. The prospect of significant share dilution, combined with the operational turbulence described in Zuckerberg’s memo, has added complexity to investor sentiment around one of the world’s most valuable technology companies.
Meta has not publicly commented on the specifics of the internal communication. However, the acknowledgment of missteps, however carefully worded, represents a notable departure from the typically measured optimism that characterises executive communications at major US technology firms.
Sources: CybernewsReuters
