Wall Street's $1.5 billion plan to build the 'McKinsey of AI'

• Anthropic has partnered with private equity giants to create a $1.5 billion "McKinsey of AI."
• The unnamed firm will build playbooks on AI transformation inside and outside PE portfolios.
• Blackstone, Hellman & Friedman, and Goldman Sachs are joined by a consortium of other investors.

Wall Street and Silicon Valley are betting that generative AI will kick off an Industrial Revolution-sized paradigm shift. And who do business leaders call when they need to make a change?

Consultants.

Anthropic is teaming up with some of Wall Street's biggest investors for a $1.5 billion joint venture for "AI-native enterprise services" — or, in easier-to-understand terms, the "McKinsey of AI," said one insider with direct knowledge of the deal.

The four founding partners, Anthropic, Blackstone, Hellman & Friedman, and Goldman Sachs' asset management arm, announced the deal Monday. The first three firms each contributed $300 million to the deal, while Goldman contributed $150 million, according to the person familiar. The remaining funding comes from a consortium of major investors, including Apollo, General Atlantic, Leonard Green, GIC, and Sequoia Capital.

"This is a rare convergence: massive market need, the unmatched AI technical capability of Anthropic, and a consortium of investors with the reach to scale fast," said Hellman & Friedman CEO Patrick Healy in a media release accompanying the announcement.

The unnamed AI firm is a way for its backers to drive returns across their portfolio companies, create a playbook for transformation through the tech, and ultimately justify the billions these firms are pouring into AI infrastructure.

"We've got 275 companies," Jon Gray, Blackstone's president, said on CNBC on Monday. "They're very interested in using Anthropic's enterprise technology, but they're saying, 'Can you help me get there? 
Can you help me change the workflow?'"

And for investors like Gray, helping them out isn't just a matter of altruism; it could materially affect their bottom lines. Labor is a $60 trillion annual cost, he said, adding: "If you can make people 15% more efficient, that's $9 trillion."

Private equity firms have been under pressure to generate returns as holding periods have been stretching. Almost 40% of companies are now held for more than 5 years, according to Bain, up from 29% in 2019.

"The demand is off the charts," said Brian Mulberry, chief market strategist at Zacks Portfolio Management, who sees the venture as the convergence of several forces: growing computing power as data centers multiply, the urgency for financial sponsors to offload aging assets, and the soaring costs affiliated with rapid token consumption.

"When computing power grows, more data centers come online," he said. "The use of this AI as a tool to try and drive productivity then becomes very scalable."

Anthropic's growing foothold on Wall Street

A central part of the venture will involve working closely with Anthropic's engineers to embed AI into their workflows and reshape internal processes. In practice, that could include introducing AI-driven agents into existing systems, allowing companies to complete tasks faster, according to a second person familiar with the deal's logistics.

Another advantage for the venture's clients will be access to the latest upgrades to Anthropic's models, said another person familiar with the deal's logistics, adding that preemptive access to these models before the public receives them could help them address mounting cybersecurity concerns — some of which are tied to the unprecedented capabilities of Anthropic's own products.

In recent weeks, Anthropic's Mythos model, an advanced LLM that has stoked concerns among US officials, has been a fixation for banks and large corporations that are trying to secure themselves against its unprecedented cyber capabilities.

Billions on the line

If the new venture can figure out how to bring businesses into the AI era, it will likely be a very valuable company: Blackstone's own portfolio companies' spending on large language models is up 15 times in the last year.

"This is a rising tide," said Thomas T. Thomas, the CEO of private-equity AI value creation firm Teragonia. His firm has built a software platform for middle-market private equity firms to apply AI insights to their businesses, targeting a market that owns trillions of dollars in companies.

It may be a big market to serve, but it's also a hard one, Thomas said, pointing to the challenge of transforming businesses holistically within the timeframe of a private equity investor.

"These companies do need to exit, or there needs to be a liquidity event in about 60 months, or faster if possible," he said. "Doing that and telling an AI value creation story in a cohesive fashion is another matter."

There's also a longer-term payoff.

Blackstone says it is the world's largest data center investor, with over $150 billion in data centers globally and a $160 billion pipeline of potential deals. The four biggest players, Amazon, Microsoft, Meta, and Google, have planned $725 billion in spending this year, which has some worried about the sheer size of the bet.

Private equity has long influenced corporate strategy, first by breaking up the midcentury corporate conglomerates into focused businesses and then by running companies with a ruthless focus toward driving returns.

If the sector can create a repeatable playbook for artificial intelligence transformation, its data center bets are likely to pay off.