Meta Pays $2 Billion for Chinese AI Startup Manus — While Announcing 20% Layoffs

Meta's Double Game: Cut and Buy Simultaneously

Meta Platforms confirmed this week that it is spending at least $2 billion to acquire Manus — a Chinese AI startup that has made a name for itself with advanced agentic AI systems. This comes alongside reports that Meta plans to lay off between 15,000 people and up to 20% of its workforce to offset enormous AI infrastructure costs.

Manus was not chosen randomly. The company, behind the viral AI agent of the same name that generated massive attention in early 2026, has built technology for autonomous AI agents capable of performing complex multi-step tasks without human intervention — exactly what Meta needs to compete with OpenAI and Anthropic.

Why Manus?

• Agentic AI technology — Manus specializes in AI that can plan, act, and complete multi-step tasks autonomously
• Alternative to internal development — Meta's own flagship project "Avocado" is delayed until May and underperforms on benchmarks
• Competitive strategy — Meta urgently needs something that can match Claude and GPT-5 on coding and reasoning tasks

The Chinese Connection

The acquisition of a Chinese AI company will likely face regulatory scrutiny — particularly in a time of escalating AI chip export restrictions and geopolitical tensions around technology. It remains to be seen whether US authorities will intervene.

Context: Meta's AI Strategy in Freefall?

Meta has had a turbulent AI period: Llama 4 Behemoth was dropped after weak internal results, Avocado is delayed, and TBD Lab — the internal superlaboratory built specifically to win the AI race — has only delivered Vibes, an AI video app. Now they are betting $2 billion on an external solution.

My take: For CIOs, this signals that even the largest tech companies acknowledge it is faster to buy AI capabilities than build them. Meta is deploying capital aggressively — while employees pay the price. Lesson: Evaluate whether your organization should build or buy key AI capabilities. In most cases, buying or partnering is faster and cheaper than building from scratch.