In short: Anthropic is in negotiations to anchor a new joint venture with Blackstone, Hellman & Friedman, and Permira that would embed Claude across private equity portfolio companies, investing roughly $200m of its own capital into a vehicle that could raise up to $1bn from buyout firms, and taking Palantir’s forward-deployed engineer model as its template.
Anthropic is in talks to invest roughly $200m in a new private-equity-backed joint venture designed to accelerate enterprise adoption of its Claude models, according to reporting from the Wall Street Journal. The proposed structure would see buyout firms, including Blackstone, Hellman & Friedman, and Permira, take equity stakes totalling approximately $1bn in the venture, which would operate as a consulting and implementation arm helping businesses integrate Claude into their operations.
No final terms have been agreed and no timeline has been announced. But the talks represent the most aggressive step Anthropic has yet taken to turn its model leadership into a distribution network, and they arrive at a moment of intensifying competition with OpenAI for the enterprise clients who will ultimately determine whether the economics of frontier AI hold together.
What the venture would actually do
The proposed joint venture is modelled, by multiple accounts, on Palantir’s forward-deployment playbook: engineers embedded inside customer organisations, driving not just adoption but workflow transformation. Rather than relying on software subscriptions alone, Anthropic would bundle model access with advisory and implementation services, the kind of hands-on work that drives the sticky, recurring revenue that AI companies need to justify their infrastructure commitments.
The 💜 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!
The strategic logic of using private equity as the distribution layer is elegant. PE firms control thousands of portfolio companies. Rather than Anthropic approaching each enterprise independently, a joint venture with Blackstone or Hellman & Friedman gives it access to those entire portfolios in one negotiation. Each buyout firm becomes, in effect, a channel partner with both a financial incentive to see Claude adopted and direct operational influence over the companies in which it is being deployed.
Blackstone already has skin in the game beyond this new venture: the firm holds approximately $1bn in Anthropic equity, having invested $200m at a $350bn valuation in February 2026 as part of Anthropic’s Series G round. That position gives Blackstone both a strategic and a financial reason to want Claude embedded as widely as possible in the corporate world, a conflict of interest, if you want to call it that, or an alignment of incentives, depending on your vantage point.
The OpenAI comparison
Anthropic is not alone in pursuing this model. OpenAI is in parallel discussions with Advent International, Bain Capital, Brookfield Asset Management, and TPG for a comparable enterprise AI venture, with total fundraising reportedly targeting approximately $4bn. The structural differences between the two pitches are telling.
OpenAI is offering private equity firms a guaranteed minimum return of 17.5%, an incentive designed to make the investment proposition simpler for LPs and investment committees that might otherwise treat an AI joint venture as too speculative. Anthropic is offering ordinary equity in the venture, with no floor on returns. That difference is both a financial signal, Anthropic is more confident in the commercial upside, or less willing to subsidise investor risk, and a cultural one: a company founded explicitly around AI safety is unlikely to be comfortable packaging its equity in instruments that prioritise investor protection over genuine shared risk.
The Axios description of the competitive dynamic cuts to the point: “It’s a whole lot faster for OpenAI and Anthropic to partner with PE firms than to approach each of their portfolio companies independently.” Enterprise AI adoption has entered a race in which distribution, not model quality alone, will determine market share. Both companies have concluded that private equity is the fastest route to scale.
Building on existing enterprise infrastructure
The new venture, if it closes, would be the third major enterprise initiative Anthropic has launched within a single quarter. In March 2026, the company committed $100m to Anthropic’s $100M Claude Partner Network, a programme anchored by Accenture, Deloitte, Cognizant, and Infosys that provides implementation support, technical architects, and co-marketing for enterprise Claude deployments. Separately, Xero has integrated Claude directly into its accounting platform in what Xero’s partnership with Anthropic brings Claude into small business finance , a model that illustrates how deeply Claude is being embedded into software products far beyond the chat interface.
By April 2026, more than 1,000 businesses are spending over $1m per year on Anthropic services on an annualised basis, up from roughly 500 two months earlier. Enterprise customers now represent approximately 80% of Anthropic’s revenue, according to reporting based on the company’s internal figures. The proposed PE venture is designed to accelerate that concentration, and to reach the segment of the enterprise market (private-equity-owned mid-market companies) that the Claude Partner Network’s large-system-integrator anchors do not typically serve.
The IPO frame
Context matters here: Anthropic is reported to be in discussions with Goldman Sachs and JPMorgan Chase about a public listing targeting October 2026, with estimates of a $60bn fundraise. At $380bn valuation, the post-money figure on the Series G, a successful IPO requires Anthropic to demonstrate not just model capability but durable, scalable enterprise revenue. A joint venture that deploys Claude across the portfolio companies of three or four major buyout firms creates both a revenue channel and a narrative: that Claude is not merely a model but an enterprise infrastructure layer.
The financial architecture that underlies all of this is worth noting. SoftBank’s $40bn bridge loan to fund its OpenAI commitment illustrated the lengths to which AI infrastructure finance has stretched; Anthropic’s own $30bn Series G was the second-largest venture funding deal in history. The PE venture adds another dimension to this picture: Anthropic is not merely raising money from PE firms as passive investors but is now proposing to build distribution vehicles with them, a shift from capitalisation to commercialisation.
What remains unresolved
The talks are ongoing and the structure is not final. Key open questions include which PE firms will ultimately participate, how governance of the joint venture will work, and whether Anthropic will retain pricing and access controls over Claude deployments made through the vehicle. That last question is not trivial: Anthropic has shown it is willing to set boundaries around how Claude is accessed, having recently moved to restrict access to Claude via certain third-party frameworks where it judged the cost and safety dynamics to be misaligned.
A joint venture that gives PE firms commercial incentives to deploy Claude as broadly as possible across their portfolios introduces a tension with that kind of granular control. Whether Anthropic can maintain its approach to model governance independent of its commercial partners while simultaneously using those partners as its primary distribution channel is the most interesting unresolved question in this story, and perhaps the most important one for anyone who cares about what responsible AI deployment actually looks like at enterprise scale.
Get the TNW newsletter
Get the most important tech news in your inbox each week.