There’s a growing conviction among investors that the biggest returns in AI won’t come from foundation models. They’ll come from whoever figures out how to actually embed AI into the businesses that run the economy. Thrive Holdings just raised $2 billion at a $12 billion valuation to test that thesis at scale.
According to TechCrunch, the round was backed by SoftBank, D1 Capital Partners, and Altimeter Capital. Thrive Holdings operates something like a private equity firm built around AI implementation. It acquires traditional businesses, mostly in fragmented, operationally complex sectors, and then rebuilds their workflows using AI tools. The New York Times first reported the news.
The firm is a spinout of Thrive Capital, one of OpenAI’s major investors. In December 2025, OpenAI took an ownership stake in Thrive Holdings and started sending employees directly into Thrive’s portfolio companies to accelerate adoption. That arrangement matters. It’s not just access to models. It’s OpenAI engineers in the building, which is a different kind of competitive edge.
That model has company. OpenAI has partnered with large private equity to launch The Deployment Company, and Anthropic is doing something similar with Ode. Both are building teams of engineers who embed inside enterprises to implement AI into workflows. What Thrive is doing fits squarely in that trend, but with an acquisition-first approach rather than a consulting or licensing model.
So far, Thrive has built two operating platforms. Current is 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 at participating firms by over 30%. Shield is its IT arm, covering around 20 companies. AI products on Shield have cut help desk resolution times by 36x, and the platform has doubled its deployed custom agents in the past month alone. Across both platforms, Thrive says it has more than 70 businesses.
The new capital will fund a third platform focused on regulatory services for the built environment. The spokesperson described it as “the work required to get physical assets approved, built, certified, and kept in operation.” Think permitting, inspection documentation, compliance tracking, and reporting across sectors like data centers, manufacturing, healthcare, power, water, and transportation.
The pitch is straightforward. U.S. infrastructure projects are often bottlenecked not by physical constraints but by regulatory complexity, and that complexity is largely manual, document-heavy, and fragmented across local jurisdictions. AI can’t replace field work or professional sign-off, but it can compress the administrative layer significantly.
- Permit preparation and research automation
- Inspection documentation and compliance tracking
- Reporting workflows across local regulatory frameworks
- Coordination across data centers, manufacturing, and public infrastructure projects
Thrive founding member Anuj Mehndiratta told TechCrunch that the U.S. needs to build and modernize critical infrastructure but that projects are frequently constrained by local, technical, and regulatory complexity. Fellow founding member Kareem Zaki added that AI combined with expert practitioners can compress regulatory bottlenecks while keeping safety standards intact and reducing cost and time to completion.
The infrastructure vertical is a smart expansion on paper. It’s large, fragmented, operationally complex, and resistant to pure software plays because the domain knowledge is so specialized. Those are exactly the conditions where Thrive’s acquisition-and-embed model has an argument. But execution at scale in physical infrastructure is a different challenge than accounting software. Regulatory environments vary by city, county, and state. And the cost of a mistake is higher.
Still, the investor appetite is clear. $2 billion at a $12 billion valuation, with SoftBank in the mix, suggests the market believes Thrive’s model is one of the more credible paths to monetizing enterprise AI adoption right now. Whether the infrastructure bet pays off will depend entirely on how well the platform generalizes beyond the cleaner workflows it has already proven in accounting and IT.
