TLDR:
- Meta plans to cut over 20% of its 79,000-person workforce, potentially eliminating around 16,000 jobs.
- Meta’s Avocado model has missed its deadline three times and now trails Google, OpenAI, and Anthropic on benchmarks.
- Meta is reportedly exploring a temporary deal to license Google’s Gemini to power its own AI products.
- Meta has committed up to $135 billion in 2026 capex and $600 billion in data center spending through 2028.
Meta layoffs are under scrutiny after Reuters confirmed plans to cut over 20 percent of the company’s roughly 79,000 employees.
About 16,000 jobs could be at risk under the reported plan. The move comes as Meta ramps up AI spending to between $115 and $135 billion in 2026.
However, the company’s own AI model has faced multiple delays. Meta is also reportedly considering licensing a competitor’s technology in the interim.
Model Delays Cast Doubt on the AI Replacement Thesis
Meta’s next-generation AI model, internally codenamed Avocado, has been delayed from March to at least May 2026. Internal benchmarks showed the model falling behind Google’s Gemini 3.0, OpenAI, and Anthropic in key areas.
Those areas include reasoning, coding, and writing performance. The delay comes at a particularly sensitive time for the company.
The delay is not a one-off event. The model has slipped three separate times from its original 2025 release target. Each delay pushes back Meta’s ability to prove that AI can handle work previously done by large teams.
Social media analyst @shanaka86 captured the tension in a widely shared post. He wrote: “Mark Zuckerberg is about to fire 16,000 humans because he believes AI can replace them. His own AI cannot replace Google’s.”
He called this contradiction “the entire story of the 2026 tech economy.” Many investors and observers have since amplified the observation online.
Meta’s previous flagship model, Llama 4 Behemoth, was never released publicly at all. Now the company is reportedly discussing a plan to license Google’s Gemini temporarily. That would mean a competitor’s model running inside Meta AI products under Meta’s own branding.
CEO Mark Zuckerberg told analysts earlier this year that he was “starting to see projects that used to require big teams now accomplished by a single very talented person.”
However, the company’s AI technology has not demonstrated that capability in competitive benchmarks. The layoffs appear to be running ahead of the technology they are supposed to depend on.
Acquisitions and Capital Commitments Add Financial Weight to the Strategy
Meta’s capital expenditure for 2026 is projected between $115 and $135 billion. That is nearly double the roughly $72 billion the company spent on infrastructure last year.
Additionally, Meta has committed to $600 billion in total data center spending through 2028. The scale of that commitment makes the AI model delays all the more consequential.
The company has also moved aggressively on acquisitions in a short time. Meta paid $14.3 billion to bring in Alexandr Wang from Scale AI.
It then spent over $2 billion on Manus and an undisclosed amount on Moltbook. Both deals came within recent months, adding to the company’s growing cost base.
The integration of these acquisitions, however, depends on a model that is still unfinished. Manus processes 147 trillion tokens using third-party AI models, not Meta’s own. Moltbook’s agent systems run on a platform called OpenClaw, also external to Meta’s stack.
Meta hired Nat Friedman, the former GitHub CEO, as part of its talent push. The company also recruited top AI researchers with compensation packages reported to exceed $100 million each. Zuckerberg described the goal as building “the highest talent density lab in the industry.”
Meta spokesperson Andy Stone pushed back on the layoff reports. He called them “speculative reporting about theoretical approaches,” with no confirmed plans or timelines.
Meta’s stock still fell 3.83 percent when the news broke. The proposed cuts would be the company’s largest since the 2022–23 efficiency drive, which removed 21,000 positions.
The post Meta’s $135 Billion AI Push Is Stumbling: Layoffs Loom as Flagship Model Trails Rivals appeared first on Blockonomi.
Source: https://blockonomi.com/metas-135-billion-ai-push-is-stumbling-layoffs-loom-as-flagship-model-trails-rivals/
