Meta cuts about 700 jobs as it shifts spending to AI

Steven

Meta cuts about 700 jobs as it shifts spending to AI

Summary

Meta has started a round of job cuts affecting roughly 700 roles as it redirects spending toward AI — building datacentres, training large language models and hiring AI talent. The reductions hit Reality Labs, parts of the social media division and recruitment teams. Meta says the changes are intended to streamline operations to better support its AI strategy announced by Mark Zuckerberg earlier this year.

Higher-level plans reported elsewhere include a potential 20% workforce reduction (about 15,000 roles) though the current cuts are far smaller. Meta ended January with about 78,800 employees. The company has ramped up costs to compete in the AI arms race: 2025 expenses rose to $118 billion and planned 2026 spending is in the $162–167 billion range, with capital expenditure of $115–135 billion targeted largely at datacentres and AI infrastructure. Meta is also developing in-house MTIA chips for GenAI inference and training, while internal model projects such as the ‘Avocado’ reasoning model have reportedly faced delays.

Key Points

  • Meta laid off about 700 employees as it prioritises AI-related investment and restructuring.
  • Job losses primarily affect Reality Labs, the social media division and recruitment teams.
  • Reuters has reported a broader plan to cut up to 20% of staff (≈15,000), though only a portion appears to have been executed so far.
  • Meta’s spending on AI and infrastructure has surged: 2025 expenses rose to $118bn; 2026 guidance targets $162–167bn with heavy capex for datacentres.
  • The company is building custom MTIA chips for GenAI workloads and has delayed some internal model releases (eg, Avocado).
  • Executive tensions and large hiring offers have marked Meta’s push for AI talent; ROI and inference scaling remain uncertain according to CFO comments.

Why should I read this?

Because this is where the money — and the risk — is going. If you follow AI, tech jobs or cloud/datacentre economics, Meta’s reshuffle signals how Big Tech prioritises compute and models over other projects (yes, even the metaverse). Short version: people get moved while servers and models get fattened.

Context and Relevance

This move matters because Meta is a major player in the AI race against OpenAI, Google and Anthropic. The cuts and huge capex plans show the company doubling down on owning compute and model development rather than outsourcing. That has implications for the labour market (hiring and layoffs), the chip and datacentre supply chains, and competition for inference capacity if AI products scale quickly. Investors and partners will watch whether these bets deliver the expected returns and whether further workforce reductions follow.

Source

Source: https://go.theregister.com/feed/www.theregister.com/2026/03/25/meta_cuts_700/