Meta Streamlines AI Department Architecture, Lays Off 600 People, Superintelligence Lab Reduced to Under 3000 Employees
Abstract
Meta announced on Wednesday, October 22, 2025 (ET), that it is laying off approximately 600 employees in its artificial intelligence division. This move aims to streamline the organizational structure, reduce management layers, and enhance operational flexibility. The layoffs primarily affect the AI infrastructure team, the Fundamental AI Research (FAIR) department, and other product-related roles, but did not impact the newly formed TBD Labs team established this summer. Following the layoffs, the headcount at Meta's Superintelligence Labs has fallen below 3,000.
Background and Reasons for Layoffs
These layoffs represent a significant step in Meta's efforts to optimize its AI division's structure. According to informed sources, Meta has long considered its AI division to be overly bloated, with research teams like the Fundamental AI Research (FAIR) department often competing with more product-oriented business units for computing resources.
Alexander Wang joined Meta in June 2025 as part of the company's $14.3 billion investment plan in Scale AI. When the newly formed "Superintelligence Labs" took over the former Meta AI division, it inherited the issue of an oversized organizational structure. These layoffs are precisely aimed at further optimizing this department's structure and solidifying Alexander Wang's leading role in the company's AI strategy.
Affected Departments and Personnel
The current round of layoffs primarily impacts the following departments:
- AI infrastructure team
- Fundamental AI Research (FAIR) department
- Other product-related roles
Notably, employees of TBD Labs were not affected by these layoffs. This lab brings together a group of top AI talent who joined Meta this summer. These team members, directly led by Alexander Wang, were spared, highlighting Meta CEO Mark Zuckerberg's preference for relying on the high-profile, highly-paid new talent he brought in.
According to sources, after this round of layoffs, the headcount at Meta's Superintelligence Labs has dropped below 3,000.
Layoff Arrangements and Compensation
Meta notified some affected employees on October 23, with their departure date set for November 21, 2025. Prior to this, employees are in a "non-working notice period." The notification states that during this period, employees' internal access will be removed, and they are not required to perform any work tasks for Meta, but they can use this time to seek other positions within Meta.
Meta will provide affected employees with 16 weeks of severance pay, plus an additional two weeks' salary for each full year of service, with the notice period deducted.
AI Strategy Adjustment
In recent months, Meta has been actively adjusting its artificial intelligence strategy, striving to keep pace with competitors like OpenAI and Google, investing billions of dollars in infrastructure projects and recruitment.
It is understood that Zuckerberg was previously dissatisfied with Meta's progress in the AI field, especially after the Llama 4 series models, released in April 2025, failed to receive positive developer feedback.
Following the substantial investment in Scale AI, Zuckerberg announced the formation of the new "Meta Superintelligence Labs," bringing together top AI researchers and engineers globally. The lab is co-led by Alexander Wang and former GitHub CEO Nat Friedman.
Financial Investment and Future Plans
During its Q2 2025 earnings call in July, Meta projected total expenses for 2025 to be between $114 billion and $118 billion, raising the lower end of its previous forecast range. The company also noted that due to ongoing AI projects, the annual expense growth rate for 2026 is expected to be higher than in 2025.
Meta is scheduled to release its Q3 financial results report next week.
Just on Tuesday, October 22, Meta also announced a $27 billion partnership with Blue Owl Capital to fund and develop the Hyperion hyperscale data center in rural Louisiana. Zuckerberg stated in a July post that the data center is expected to be large enough to cover "a significant portion of Manhattan."
Industry Context
These layoffs are part of a broader trend in the tech industry. Driven by economic uncertainty and emerging technologies like AI, multiple tech giants are undergoing organizational restructuring and resource reallocation. According to data from Layoffs.fyi, several tech companies have conducted layoffs in 2025 to date, with total layoffs exceeding 12,000 people.
Meta's latest layoffs indicate that even while heavily investing in AI, tech companies still need to optimize organizational structures and improve efficiency to navigate intense market competition. This move also reflects Meta's clear direction in its AI strategy: focusing on core teams, reducing redundancy, and accelerating the pace of innovation.