OpenAI-backed Thrive Holdings raises $2 billion to wire AI into everyday businesses
Thrive Holdings buys traditional firms, sends in AI experts, and now wants to tackle the mountain of paperwork that slows down building roads, data centres, and power plants.

Key points
- Thrive Holdings raised $2 billion at a $12 billion valuation, with backing from SoftBank, D1 Capital Partners, and Altimeter Capital.
- OpenAI took an ownership stake in Thrive Holdings in December 2025 and has sent employees to help its companies adopt AI tools.
- Thrive's accounting arm, Current, processed more than 7,000 tax returns using AI at 98% accuracy, cutting prep times by over 30%.
- A new third platform will target regulatory services, the permitting, inspection, and compliance work needed to get physical infrastructure built and running.
- Thrive now has more than 70 businesses on its platform, spread across accounting and IT services.
Thrive Holdings has secured $2 billion in fresh funding, first reported by TechCrunch, and the company's pitch is refreshingly unglamorous: buy ordinary businesses, plug AI into their daily work, and watch them get faster and cheaper.
Think of it as a private equity firm, a type of investment company that buys businesses and tries to improve them, but with AI as the main tool. Thrive already owns more than 70 companies across two arms. Current, its accounting division, has over 50 firms and 2,000 professionals. Shield, its IT services arm, covers around 20 companies.
What is the OpenAI connection?
Thrive Holdings spun out of Thrive Capital, one of OpenAI's biggest early investors. In December 2025, OpenAI took a direct ownership stake in Thrive Holdings and began sending its own staff to work inside Thrive's companies.
That arrangement is the real story here. OpenAI and Anthropic, two of the biggest AI labs, have both moved into this kind of hands-on corporate AI work. OpenAI helped launch a venture called The Deployment Company; Anthropic backed one called Ode. Both models involve teams of engineers embedding inside businesses and rebuilding how they operate.
Does any of this actually work?
The numbers Thrive published suggest yes, at least so far. Current's tax AI product, called TaxAI, processed more than 7,000 returns at 98% accuracy and cut the time accountants spend on each return by more than 30%. Shield's AI tools have cut IT help-desk resolution times by a factor of 36.
For the staff at those firms, that means less time on repetitive paperwork. It does not, Thrive says, mean replacing people.
| Platform | Focus | Scale | Headline result |
|---|---|---|---|
| Current | Accounting | 50+ firms, 2,000+ staff | Tax prep time down 30%+ |
| Shield | IT services | ~20 companies | Help-desk speed up 36x |
| New (unnamed) | Regulatory services | Launching now | Target: physical infrastructure |
What is the new platform about?
Part of this funding opens a third area: the regulatory maze around physical infrastructure. Building a data centre, a hospital, or a water treatment plant involves stacks of permits, inspections, compliance reports, and sign-off from multiple agencies. Thrive wants to use AI to handle the research, document preparation, and tracking, while qualified professionals still make the actual decisions.
As founding member Anuj Mehndiratta put it: the US needs to build more critical infrastructure, but projects keep getting stuck in local and regulatory complexity.
AI will not replace the site inspector or the licensed engineer. It can, though, draft the permit application faster.
Common questions
Does this affect me if I use an accounting firm?
Possibly, yes. If your accountant is part of Thrive's Current network, AI may already be handling parts of your tax return. Thrive says accuracy is high, but it is worth asking your firm how they use AI tools and who reviews the final figures.
Is my financial data safe with an AI accounting platform?
That is the right question to ask. Thrive has not published detailed public documentation on data handling. Before any firm uses AI on your returns, ask them directly which AI tools they use, where your data is stored, and whether it is used to train any AI models.



