Alibaba ended 2025 with 128,197 employees. A year earlier, it had 194,320. That’s 66,000 jobs gone—a 34% reduction that makes this one of the largest workforce contractions in tech history.
The same week it disclosed those numbers, Alibaba launched Wukong, an enterprise AI agent platform, and announced price increases of up to 34% on AI chips and cloud storage. The message is clear: fewer people, more automation, higher margins.
What Alibaba Actually Did
Not all 66,000 departures were layoffs. A significant portion came from divesting Sun Art (hypermarkets) and Intime (department stores)—labor-intensive retail businesses that didn’t fit the AI-first vision. These were asset sales, not pink slips.
But the restructuring goes deeper. Alibaba pulled its core AI research, enterprise services like Wukong, and consumer applications under one roof. The reorganization signals that everything outside AI and cloud is now secondary.
CEO Eddie Wu laid out the ambition during the earnings call: $100 billion in annual AI and cloud revenue over the next five years. That’s roughly quadruple current levels.
The Wukong Play
Wukong isn’t just another chatbot wrapper. It’s an agentic AI platform designed to coordinate multiple AI agents that automate enterprise workflows—the kind of work that previously required teams of people.
Alibaba launched it March 17, directly targeting commercial customers who want AI that does things, not just AI that suggests things. The timing alongside the price hikes is intentional: Alibaba is testing whether enterprise customers will pay premium prices for agents that replace human coordinators.
The price increases themselves are aggressive. T-Head AI chips (Alibaba’s custom silicon) went up by mid-30% range. Cloud storage fees rose about 30%. This isn’t inflation adjustment—it’s a bet that demand is inelastic enough to absorb serious markups.
The Strategic Calculation
Follow the math. If AI agents can handle work that previously required three or four employees, and you can charge 34% more for the infrastructure running those agents, you’ve fundamentally changed the business model. Fewer people on your payroll, more revenue per AI workload, higher margins overall.
Alibaba is running the same playbook Western companies are running, just faster and more explicitly. Block laid off 40% of its workforce citing AI. Meta’s cutting staff while spending $135 billion on AI infrastructure. Oracle may eliminate 30,000 positions to fund data centers.
The difference is scale. No Western company has shed a third of its workforce in twelve months while simultaneously launching an enterprise AI product designed to accelerate the same trend at customer companies.
Who Wins, Who Loses
Winners:
- Alibaba shareholders, if the $100B target materializes
- Enterprise customers who can automate workflows through Wukong
- Remaining Alibaba employees, who are presumably positioned for AI-centric roles
Losers:
- 66,000 former Alibaba employees (though many transferred with divested assets)
- Chinese tech workers watching their industry consolidate around AI
- Competitors who need to match both the AI investment and the price increases
The uncertain middle:
- Enterprise IT buyers, who now face higher cloud costs alongside pressure to automate
- Alibaba’s cloud competitors, who must decide whether to follow the price hikes or sacrifice margins
The Broader Signal
Alibaba’s transformation is the most aggressive example yet of a major tech company restructuring itself entirely around AI economics. It’s not hedging. It’s not running pilot programs. It’s cutting a third of its workforce, hiking prices, and launching an agent platform all in the same quarter.
This is what conviction looks like. Whether it’s correct conviction remains to be seen.