Meta Cuts 10% of Workforce to Fund AI

Meta is eliminating about 8,000 employees, roughly 10% of its global workforce, and will close approximately 6,000 open positions, with the first wave of cuts targeted for May 20. The company frames the action as a tradeoff to run the business more efficiently and to offset large-scale spending on artificial intelligence, including a $115B to $135B capital expenditure guidance for 2026, up from $72.2B in 2025. Janelle Gale, Meta's chief people officer, wrote, "We're doing this as part of our continued effort to run the company more efficiently and to allow us to offset the other investments we're making."

The reorganization is explicitly tied to concentrating engineering and product resources around AI workstreams. Engineers have been moved into a centralized Applied AI Engineering organization focused on building autonomous coding agents and AI-native tooling. Prior cuts in Reality Labs and studio shutdowns reduce headcount in lower-priority hardware and metaverse efforts. Reported severance provisions for U.S. employees include 18 months of COBRA health insurance coverage and a base pay severance of 16 weeks, increasing by 2 weeks per year of service; international packages will vary by country.

Expect hiring freezes or role eliminations in non-AI product lines, redeployment of ML engineers and infra teams into model development and productionization pipelines, and accelerated investment in compute, data platforms, and MLOps. The company's capital guidance implies significant procurement of cloud and on-prem GPU/accelerator capacity, increased model training budgets, and expanded annotation and evaluation workloads.

This is not a cost-driven distress layoff; Meta remains highly profitable and generated record quarterly revenue in late 2025. The cuts are a strategic reallocation to prioritize AI capabilities and products over legacy initiatives. Meta's scale means these choices will reshape hiring markets for ML engineers, prompt competitors to adjust talent strategies, and increase commercial demand for AI infrastructure and services. The move parallels broader industry behavior where firms trade headcount for automation and AI investment, but at Meta's scale the downstream effects on open roles, vendor demand, and partner ecosystems will be sizable.

Centralizing talent into Applied AI Engineering and similar AI-first units accelerates model-focused product timelines and raises the internal bar for compute, tooling, and production ML practices. Practitioners should anticipate new internal templates for model governance, reproducible training, and end-to-end MLOps as Meta scales experiments into consumer-facing agents. The severance and redeployment terms also signal the company expects to reuse institutional knowledge where possible, but will not hesitate to cut redundant or lower-priority workstreams.

With Meta committing to very large capital spending and a concentrated engineering effort, expect faster iteration on large multimodal models, chatbots, and coding agents. This changes competitive dynamics with OpenAI, Anthropic, and major cloud providers, and will increase demand for third-party tooling that speeds model development and reduces training costs.

Short-term: Meta's Q1 2026 earnings call and guidance update on April 29 will clarify how pockets of revenue and margins absorb the AI spending. Medium-term: hiring patterns in ML, ML infrastructure vendor procurement signals, and announcements about new AI products, partnerships, or internal governance frameworks. The possibility of additional layoff rounds later in 2026 remains an open variable as Meta adjusts to AI progress and efficiency gains.