Meta Begins 8,000 May 2026 Layoffs to Fund a $145B AI Bet

Meta will begin cutting roughly 8,000 jobs the week of May 20, 2026, and has frozen another 6,000 open roles, according to CNBC reporting on the rollout. The reductions land across Reality Labs, the Facebook product organization, recruiting, sales, and global operations, and they arrive in the same quarter Meta posted $56.3 billion in revenue and $26.8 billion in net income. The contradiction is the story. This is a profitable company laying off about one in every ten employees while raising its 2026 AI capital expenditure guidance to a range of $125 billion to $145 billion. Zuckerberg is not cutting because Meta is shrinking. He is cutting because he is rebuilding the company around AI infrastructure, and everything outside that thesis is being treated as overhead. The May wave is the largest single layoff action Meta has taken since the 2023 "Year of Efficiency," and unlike that round it is being executed during a quarter of record profit.

What Meta Is Actually Cutting in May 2026

The headline number is about 8,000 layoffs starting the week of May 20, but the structural number is closer to 14,000 once the 6,000 cancelled open requisitions are included. The cuts are not concentrated in a single org. CNBC's reporting and follow-on coverage describe impact across Reality Labs (the VR/AR group that builds Quest, Ray-Ban Meta, and Horizon Worlds), the Facebook social division, the recruiting function, sales, and global operations. Middle management and non-AI product roles are reportedly absorbing a heavier share of the cuts than direct AI research teams. Within Reality Labs the May cuts compound a January 2026 round that already trimmed roughly 10% of that division's headcount; the unit has gone from a flagship priority to a tightly managed cost center in less than 18 months. Inside the Facebook social org, the cuts are reportedly weighted toward managers of managers and toward product surfaces that no longer move daily active usage at a level Zuckerberg considers material. Global operations is absorbing a wave of consolidation in customer support, trust and safety, and policy roles that the company is increasingly routing through Llama-powered internal tooling.

This is not Meta's first 2026 reduction. Earlier in the year, the company cut roughly 1,000 to 1,500 Reality Labs employees, about 10% of that division, and trimmed budget there by an estimated 30%. Performance-based reductions ran through Q1, with several thousand workers exited under the company's "low performer" framework that has been a recurring feature of every Meta cycle since 2023. Stock grant portions were cut by 5% in February, on top of a 10% reduction the prior year. Median compensation across the company slipped from $417,400 in 2024 to $388,200 in 2025, per Meta's own filings. The May wave is the largest single action of the year so far, and Meta is telegraphing more: a potential August round and additional cuts later in 2026, with the timing and scope still being finalized. Internal Blind threads in the lead-up to May 20 describe workers gaming PSC ratings, hoarding documentation of their own contributions, and trading rumors about which orgs will be next. The same threads surface a recurring complaint about internal monitoring tools that track keystrokes and mouse movement under the Model Capability Initiative deployed in April.

Severance specifics for the May cohort have not been disclosed in detail. Past Meta layoffs in 2022 and 2023 came with 16 weeks of base pay plus two weeks per year of service, accelerated vesting of any RSUs scheduled to vest within the next 15 days, COBRA continuation typically covered through six months, and outplacement support via firms like LHH and Lee Hecht Harrison, but workers should not assume the same package this round until termination paperwork is in hand. The 15-day RSU acceleration window is the single highest-stakes line item for senior engineers and managers because the difference between vesting on the Friday before separation and the Monday after has historically been five to six figures of equity. WARN Act notification thresholds apply at the state level, and California, New York, and Washington all have stricter triggers than the federal 60-day baseline, which is why Meta sequences cuts in those states with longer notice periods or pay in lieu of notice. The pattern of companies quietly cutting benefits alongside layoffs is one reason to read the fine print closely. For broader context on how Meta employees themselves are framing this round internally, see our earlier coverage of the Blind threads describing morale inside the company and the nine severance moves workers should make in the first week.

Why This Round Feels Different From the 2023 "Year of Efficiency"

Meta's 2023 layoffs cut roughly 21,000 jobs across two waves and were sold as a correction to pandemic-era over-hiring. The narrative was tidy. Too many people were hired in 2021 and 2022, and the company needed to right-size to grow margins. That story does not fit the 2026 cuts. Meta is hiring aggressively in AI research, reportedly offering compensation packages reaching $100 million for senior researchers joining Meta Superintelligence Labs, and it is committing tens of billions of dollars to AI compute. The 2026 reductions are a reallocation, not a retrenchment. The 2023 cohort was told the company was over-hired; the 2026 cohort is being told the company is under-investing in the wrong people.

The capex numbers tell the story. Meta spent $39.2 billion on capital expenditures in 2024 and $72.2 billion in 2025. The 2026 guidance now sits between $125 billion and $145 billion, a jump that includes a recent $21 billion commitment to cloud provider CoreWeave and a $27 billion joint venture with Nebius for a gigawatt-scale AI data center in Louisiana. In Q1 2026 alone, Meta booked $107 billion in new cloud and infrastructure contractual commitments. Every dollar redirected into GPUs, fiber, and custom silicon has to come from somewhere, and the somewhere is increasingly people, particularly people in support functions and non-AI product lines.

This is the same pattern visible across the industry. LinkedIn's parent Microsoft cut about 5% of its workforce earlier in 2026 while raising its own AI spend. Walmart cut 1,000 corporate roles without citing AI directly. Coinbase trimmed 700 in an AI pivot. Snap cut roughly 1,000 jobs in April 2026 under similar AI-funding logic, and Pinterest's 2026 AI-driven reduction cleared the way for a parallel infrastructure push at a much smaller scale. The shared logic is that AI infrastructure is becoming the dominant line item, and headcount is the variable being squeezed to fund it. Our earlier analysis of Zuckerberg's broader AI capex thesis maps the financial mechanics in more detail, and the corporate euphemism decoder catalogs the language companies use when restructurings are really AI-funding moves. The blunt version: in 2023 the cuts paid for a higher stock price, and in 2026 they pay for GPUs.

There is also a structural mismatch that the 2023 cuts did not have. Meta is competing with xAI, OpenAI, Anthropic, and Google for a thin layer of senior AI researchers (the xAI talent exodus has pulled directly from Meta and back) while simultaneously shrinking the support layer around them. Recruiters are being cut even as the company tries to close $100 million offers. Sales teams are being cut even as ad revenue funds the AI build. That tension is the "reality check" in this round: the company that wants to win superintelligence is the same company asking remaining staff to do more with fewer colleagues. A staff engineer who shipped ranking improvements on Reels in 2023 was a clear keeper; the same engineer in 2026 needs to demonstrate AI-adjacent infrastructure work or a direct revenue line, because Reality Labs proximity and pure consumer product work are no longer protected categories.

What Meta Survivors and Job-Seekers Should Do Now

If you were cut: read your severance paperwork before signing anything, particularly any release tied to extended healthcare or stock vesting. Meta's prior packages have been generous by industry standard, but no two waves are identical, and the company has been trimming compensation at the edges. File for unemployment in your state the day your last paycheck lands. Eligibility windows are tighter than most people expect, and California specifically has rolled back several pandemic-era extensions. If you held unvested RSUs, ask explicitly about accelerated vesting and the exact cutoff date. The difference between a Friday and a Monday termination has historically been five figures for senior engineers. Confirm whether your separation date triggers WARN Act pay in lieu of notice in your state, request the severance agreement in writing, and do not sign in the 24 hours after the call; most agreements give you seven to 21 days for a reason.

If you survived the cut: assume another round is coming. Meta has publicly signaled August and late-2026 cuts. Update your resume and LinkedIn now, not in August. Internal mobility is the cleanest hedge. AI infrastructure, applied research, ranking and recommendations, and the small set of teams directly attached to Llama and Meta Superintelligence Labs are the parts of the company still hiring. If you sit in a Reality Labs role that has survived three rounds of cuts, build a backup story for why your work feeds AI or core revenue, because the budget conversation is being held in those terms. Document your wins in the language of capex efficiency or revenue per employee, since those are the metrics being used to scope the August round. Take any rotation into an AI-adjacent team, even at lateral comp.

If you are looking externally: the highest-leverage skill in this market is the ability to ship product on top of foundation models, not the ability to train them. Former Meta engineers and researchers from the past two years have landed disproportionately at xAI, Anthropic, OpenAI, Google DeepMind, Snowflake, and a long tail of AI infrastructure startups, which is where the open-headcount budget has migrated. Forward deployed engineers are the most in-demand role of 2026, Google's cloud AI deployment teams are hiring aggressively, and the coder-to-AI-manager track is where former Meta product engineers are landing. Recruiter-side, expect the search to be slower than 2022. Recruiters are now averaging more than 300 applications per role. Apply early in the week, target roles posted in the last 72 hours, and treat referrals as the only reliable signal. The broader worker playbook for staying valuable as AI reshapes the job is worth a re-read this week.

A practical note on industry posture: do not assume Meta cuts mean tech is closing. The same week of layoffs, Meta confirmed $107 billion in new cloud commitments. The money is moving from one column to another, not leaving the sector. Smaller social and ad platforms are quietly paying out retention bonuses to lock in senior engineers they expect Meta and the AI labs to court next. Our broader read on the US tech sector layoff wave and why America is unprepared for the AI layoff cycle put the May cuts in a wider frame, and the entry-level squeeze created by AI raising the productivity bar explains why this round is hitting recent graduates and early-career workers hardest.

People Also Asked

Q: How many people is Meta laying off in May 2026?

A: Roughly 8,000 employees, or about 10% of Meta's global workforce, starting the week of May 20, 2026. The company is also cancelling around 6,000 open job requisitions, bringing the effective headcount reduction closer to 14,000 slots. Additional rounds are expected in August and later in 2026.

Q: Which Meta divisions are most affected by the May 2026 cuts?

A: Reality Labs, the Facebook social division, recruiting, sales, and global operations are all reported to be hit, with middle management and non-AI product roles absorbing a heavier share. AI infrastructure, applied research, and the Meta Superintelligence Labs group remain in active hiring mode.

Q: Why is Meta cutting jobs if it just posted record revenue?

A: Q1 2026 revenue hit $56.3 billion with $26.8 billion in net income, but Meta has also raised its 2026 capital expenditure guidance to $125 billion to $145 billion, largely to fund AI infrastructure including a $21 billion CoreWeave commitment and a $27 billion Louisiana data center joint venture with Nebius. The layoffs free operating budget to feed that capex line.

Looking for your next opportunity? Join Metaintro and let our AI scout matching roles in the parts of the market that are still hiring, including the AI infrastructure, applied research, and forward-deployed engineering teams that survived this round of cuts.