Opendoor just shut down its India operations less than two years after opening offices there, and the move has touched a nerve across Silicon Valley. As TechCrunch AI reports, the San Francisco home-buying platform is pulling operational work back to the U.S. and shifting toward smaller, AI-native teams. CEO Kaz Nejatian framed it as a decision about where customers are. Investors and outsourcing analysts heard something bigger: an early signal that AI is rewriting the math behind offshore work.
What stands out here is the timing and the language, not just the headcount.
Why this one closure got attention
India isn’t a back-office afterthought anymore. It’s the world’s largest Global Capability Center market, with more than 2,100 centers employing about 2.36 million people and generating close to $100 billion a year, according to TechCrunch AI. So when a U.S. company unwinds an India team and openly credits AI-native, leaner workflows, people in that ecosystem pay attention.
Nejatian said Opendoor had built a large India team to handle manual work across fragmented systems. That’s exactly the kind of labor automation targets first.
The honest complication
Here’s where the story gets messy, and it matters. Opendoor has been cutting across the whole business. Securities filings show it employed 1,042 people globally at the end of last year, down from 1,470 a year earlier. Its non-U.S. workforce fell to 184 from 342 over the same stretch. The U.S. housing market hammered online home-buyers, and Opendoor has been slashing costs everywhere.
So this isn’t a clean before-and-after on AI. It’s a struggling company trimming everything, with AI language layered on top. Worth holding both ideas at once.
What the experts actually said
The commentary in the TechCrunch AI piece splits into two camps:
- The job-loss read. Sheel Mohnot of Better Tomorrow Ventures put it bluntly: “As manual work gets replaced by AI, a lot of jobs will be lost in India.” Keshav Lohia at Emergent Ventures called it a “watershed moment” for AI-driven operations, arguing AI is starting to break the cost-arbitrage model that built India’s offshoring industry.
- The structural read. Phil Fersht, CEO of HFS Research, which tracks the outsourcing industry, pushed back on the simple “jobs moving to the U.S.” framing. The real shift, he said, is that AI cuts how much operational labor companies need at all, letting firms run leaner no matter where they sit. “This is not an isolated restructuring,” Fersht said. “It is part of a much broader pattern.”
Fersht’s term for the winning model is worth remembering: Services-as-Software. Companies that fuse AI, software, and human expertise to deliver outcomes without piling on headcount. He expects Opendoor to be one of the first high-profile examples, not the last.
Where this goes next
Project this out one to three years and the direction is clear, even if Opendoor itself is a noisy data point.
- The arbitrage moat shrinks. Cheap labor was the pitch. When AI handles the repetitive workflow layer, paying for distance and lower wages stops being the obvious win. Varun Rekhi of Speedinvest warned this could eventually pressure one of India’s most important export industries.
- “Leaner” becomes the default org design. Expect more founders to brag about revenue-per-employee and AI-native teams. That language is already a status marker among investors.
- The work changes shape, it doesn’t vanish overnight. The centers that survive will move up the value chain into R&D, judgment, and AI oversight. Pure manual processing is the exposed layer.
What to do about it
If you run operations or a services business, audit which of your workflows are “manual work across fragmented systems.” That’s the phrase Nejatian used, and it’s the bullseye for automation. Move your people toward judgment, exceptions, and outcomes that AI can’t close on its own.
If you’re building, the Services-as-Software model is the one to study. Sell the result, not the seat count.
And if you’re reading the headlines, keep the skepticism Opendoor itself demands. A company cutting jobs across the board is not clean proof of an AI revolution. It’s a real trend wearing a complicated costume.
One closure doesn’t settle the debate. But the people who watch outsourcing for a living are no longer treating this as a question of if. You can find the full reporting at the original TechCrunch AI source.