OpenAI-backed Thrive Holdings just raised $2 billion at a $12 billion valuation, and the money is aimed squarely at dragging traditional businesses into the AI era. According to TechCrunch AI, the round drew in SoftBank, D1 Capital Partners, and Altimeter Capital, with the New York Times first to report the news. This isn’t a chatbot startup or a model lab. It’s something stranger, and arguably more telling about where AI is actually making money.
What Thrive actually does
Think of Thrive Holdings as a private equity firm with an AI twist. Instead of just buying companies and cutting costs, it buys traditional businesses like accounting firms and IT shops, then rebuilds their workflows around AI agents.
The firm runs on two pillars so far, per TechCrunch AI:
- Current, its accounting arm, with more than 50 firms and over 2,000 professionals. Its self-improving tax agents, called TaxAI, processed more than 7,000 returns at 98% accuracy and cut tax prep times by over 30%.
- Shield, its IT arm, with around 20 companies on the platform. Its AI products sped up help desk resolution times by 36x, and it doubled the number of custom AI agents deployed in just the last month.
Across both platforms, Thrive now counts more than 70 businesses. Those numbers are the pitch. Investors aren’t betting on a demo. They’re betting on returns that already showed up in real firms.
The OpenAI connection
Here’s what makes this more than another PE raise. Thrive Holdings is a spinout of Thrive Capital, one of OpenAI’s major backers. In December 2025, OpenAI took an ownership stake in Thrive Holdings, and part of that deal involved OpenAI sending its own employees to work inside Thrive’s companies to speed up AI adoption.
That hands-on model has quietly become a business of its own. And it helps explain the investor enthusiasm behind this raise.
Thrive isn’t alone in this play. OpenAI and Anthropic have both teamed up with large private equity firms to launch similar ventures, The Deployment Company and Ode, respectively. Both are billion-dollar efforts building teams of elite engineers who embed inside enterprises and wire AI into their workflows.
Why this matters
What stands out here is the shift in strategy. The AI industry spent the last few years selling tools and hoping companies would figure out how to use them. Adoption lagged. Turns out most businesses don’t want a model. They want the outcome.
So the labs changed tactics. Instead of shipping software and walking away, they’re buying the business, staffing it with their own engineers, and proving the value line by line. It’s slower, it’s expensive, and it’s a lot more hands-on. It also sidesteps the biggest problem in enterprise AI: nobody knew how to actually deploy it.
This is significant because it reframes what an AI company can be. Thrive looks less like a software vendor and more like an operator that happens to run on AI.
What comes next
Part of Wednesday’s raise will fund a third platform, this one focused on regulatory services for the built environment. A spokesperson described it as “the work required to get physical assets approved, built, certified, and kept in operation.”
The logic is that infrastructure projects choke on complexity. “The U.S. needs to build and modernize more critical infrastructure, but projects are often constrained by local, technical, and regulatory complexity,” founding member Anuj Mehndiratta told TechCrunch. That covers data centers, manufacturing, healthcare, power, water, and transportation.
Mehndiratta is clear that AI won’t replace field work, local judgment, or professional sign-off. What it can do is ease the manual grind: research, reporting, permit prep, inspection documentation, and compliance tracking. Fellow founding member Kareem Zaki put it plainly, saying AI paired with experts can “compress” regulatory bottlenecks while keeping safety standards high and costs lower.
Watch this model closely. If Thrive keeps posting numbers like a 36x speedup and 98% accuracy, expect more labs to stop selling AI and start buying the companies that use it. You can read the full details at the original source.