Lead: Sources told Reuters that Meta is preparing large-scale staff cuts that could affect 20% or more of its workforce as the company tries to balance soaring artificial-intelligence infrastructure costs with efficiency gains from AI-assisted work. No timetable or final percentage has been announced, and senior leaders were reportedly asked to begin planning reductions. The company employed nearly 79,000 people as of 31 December. Meta did not immediately comment on the reported planning.
Key takeaways
- Sources told Reuters the planned reductions could reach 20% or more of Meta’s workforce, though no date or final scale has been confirmed.
- Meta’s headcount stood at roughly 79,000 as of 31 December, making a 20% cut equivalent to about 15,800 roles.
- If adopted, this would be the largest staff reduction since the company’s 2022–2023 restructuring that removed roughly 21,000 roles over two rounds.
- Meta has pledged heavy AI-related capital spending, including plans to invest about $600 billion in data centers through 2028.
- The company has made large recruitment offers to top AI researchers and pursued acquisitions such as Moltbook and the reported near-$2 billion purchase of Manus.
- Recent product setbacks—notably problems with Llama 4 variants and the abandoned ‘‘Behemoth’’ release—have complicated Meta’s AI rollout.
- Other major US tech firms have announced significant cuts this year: Amazon cut ~16,000 roles (~10%) in January; Block reduced staff by nearly half in February.
Background
Over the past year Meta has intensified its push into generative AI, recruiting senior researchers with exceptionally large compensation packages and creating a superintelligence unit tasked with restoring its competitive lead. The company has announced an ambitious plan to spend roughly $600 billion to expand data-center capacity through 2028, a commitment that raises near-term capital needs and operational costs. Those investments sit alongside a stated expectation that AI will enable substantial efficiency gains, a claim CEO Mark Zuckerberg highlighted in January when he said he had started to see “projects that used to require big teams now be accomplished by a single very talented person.”
Meta previously carried out major workforce reductions in late 2022 and early 2023—about 11,000 jobs in November 2022 and another roughly 10,000 roles months later—under a restructuring billed as a “year of efficiency.” The size of the company today (near 79,000 employees) and the scale of planned AI spending have created pressure on management to reconcile large capital outlays with operating margins. At the same time, the company’s Llama 4 series encountered criticism for benchmark reporting and some internal decisions—such as shelving the largest Llama 4 variant dubbed “Behemoth”—have delayed product momentum.
Main event
According to three people who spoke to Reuters, senior Meta executives have signalled their intention to pursue substantial headcount reductions and told divisional leaders to prepare plans for cutting roles. Those sources, who requested anonymity because they were not authorised to speak publicly, said no final decision or date has been set. The reporting frames the move as a response to the dual pressures of heavy AI infrastructure commitments and the promise of productivity gains from AI tools.
Internally, managers have reportedly been asked to model where reductions could be made and how to reorganise teams to exploit AI-driven workflows. The superintelligence team—tasked with building advanced models such as the in-development “Avocado”—has been under scrutiny after Avocado’s performance did not meet internal expectations and after earlier Llama 4 issues. Parallel acquisitions this year, including social platform Moltbook and the reported purchase of Chinese AI startup Manus for at least $2 billion, illustrate Meta’s willingness to buy capabilities even as it rethinks staffing.
Meta has not issued a public statement about the planning reported by Reuters. When asked in January about efficiency gains, Zuckerberg pointed to early examples of smaller teams completing work that previously required many more people; that line of argument appears to underlie management’s current modelling. Any large-scale job cuts would be the most significant since the two rounds of reductions in 2022–2023 and would reshape organisational priorities across AI, product and operations teams.
Analysis & implications
Financially, a large reduction in headcount would reduce near-term operating expense and could help fund the company’s heavy capital programme for data centres. Meta’s plan to spend about $600 billion through 2028 on infrastructure implies ongoing capital intensity; lower payroll costs would be one lever to preserve cash and maintain investment in servers, networking and facilities. But layoffs are a blunt instrument—short-term savings must be weighed against the potential loss of institutional knowledge and slower product development in the medium term.
Operationally, the move signals a bet that AI will deliver meaningful productivity gains fast enough to offset cuts. That trade-off depends on model performance (for example Avocado and successors), the speed of internal tooling adoption, and the ability of remaining teams to scale output. If AI systems underperform or require larger teams for model tuning, the assumed efficiency dividend may not materialise, creating a gap between spending commitments and achievable outcomes.
There are reputational and regulatory risks as well. Large, sudden reductions attract public scrutiny and can trigger local labour and compliance issues across jurisdictions where Meta operates. For a company that also touts responsible AI and invests in safety research, workforce disruptions complicate messaging about long-term commitments to users and partners. Investors will watch closely for guidance on cost savings, severance commitments and the effect on product road maps.
Comparison & data
| Company | Announced cuts | Approx. % of workforce | Date (reported) |
|---|---|---|---|
| Meta (planning) | Potential, not final | Up to 20% or more | Reported Mar 2026 |
| Amazon | 16,000 jobs | ~10% | Jan 2026 |
| Block | Reduced nearly half | ~50% | Feb 2026 |
The table places the reported Meta plan in the context of large US tech workforce reductions this year. Meta’s possible 20% cut would be materially larger than Amazon’s January reductions (~10%) and smaller than the dramatic, company-specific reductions at Block (~50%). Differences reflect firm-specific business models, capital strategies and management choices about where to prioritise AI investment.
Reactions & quotes
Public comment from Meta was not available at the time of reporting; the company has declined to confirm the planning reported to Reuters. Analysts and observers have framed the reported planning as part of a wider industry shift that balances aggressive AI investment with headcount rationalisation.
“could affect 20% or more of the company”
Reuters (news agency, sourced reporting)
In public remarks earlier this year, Meta’s chief executive explicitly linked AI investments to efficiency gains, which company leaders now appear to be translating into organisational planning.
“projects that used to require big teams now be accomplished by a single very talented person”
Mark Zuckerberg (CEO, Meta)
Unconfirmed
- Whether Meta will finalise a 20% target (reports indicate planning only; no final decision or date given).
- Which specific divisions, regions or roles would be most affected if cuts proceed.
- The precise timeline for any announced layoffs and the details of severance or transition support.
- How directly recent acquisitions (Moltbook, Manus) relate to internal staffing decisions.
Bottom line
The reported planning at Meta captures a tension facing major tech firms: how to fund massive AI infrastructure while trying to extract operational efficiency from the same technology. If Meta proceeds with cuts near 20%, the company will materially reshape its workforce and accelerate a broader industry realignment around AI-driven productivity.
Key indicators to watch in the coming weeks are any official announcement with scale and timing, leadership guidance on where reductions will fall, and updates on the performance and deployment timelines for internal AI models such as Avocado. Investors, employees and partners will be attentive to how Meta balances capital spending, product development and human capital as it pursues AI at scale.