Anthropic just took a $1.5B shot at McKinsey. OpenAI did the same thing the same day.

Anthropic just announced a new joint venture with Blackstone, Hellman & Friedman, and Goldman Sachs — a $1.5 billion AI services firm whose entire purpose is to put Claude inside the operations of mid-sized companies. Anthropic, Blackstone, and Hellman & Friedman each committed $300 million. Goldman Sachs is a founding partner. The backer list includes Apollo Global Management, General Atlantic, Leonard Green, GIC, and Sequoia Capital. The headline number is $1.5B in committed capital. The headline that matters is what they are spending it on.

The same day — hours earlier, in fact — OpenAI launched a parallel joint venture of its own. It is called "The Development Company." It is raising $4 billion at a $10 billion valuation. It has 19 investors. The named ones are TPG, Brookfield Asset Management, Advent, and Bain Capital. There is zero investor overlap with Anthropic's vehicle. Two frontier labs, on the same day, built two parallel enterprise services stacks pointed at the same customer. The customer is the PE-owned mid-market.

The numbers underneath the move are the part that should make every consulting partner uncomfortable. CNBC's report cites the ratio that explains the whole thing: for every dollar companies spend on software, they spend six on services. The frontier labs were previously selling the dollar. They just announced they are also selling the six. The consulting industry — McKinsey, Accenture, Deloitte, BCG, and the long tail under them — is a multitrillion-dollar TAM that the labs were not addressing. They are now. CNBC also reports that 85% of PE-backed CFOs already factor AI-enabled finance capabilities into the valuations they pay. The buyer is ready.

The backstory is what makes the move sharp instead of opportunistic. Anthropic already had a Claude Partner Network. Accenture, Deloitte, and PwC are inside it. The new joint venture also joins that network — and competes with it. Anthropic's CFO Krishna Rao said it directly: "Enterprise demand for Claude is significantly outpacing any single delivery model." Translation: the existing partners are not deploying fast enough, so we are building a faster one. Blackstone's COO Jon Gray framed the same problem from the buyer side: the venture targets "one of the most significant bottlenecks to enterprise AI adoption." Goldman Sachs's Marc Nachmann said the goal is to "democratize access to forward-deployed engineers." Forward-deployed engineers — FDEs — are the structural element to watch. Palantir built a $250 billion company on that model. Anthropic and OpenAI both just adopted it.

The countervailing angle is the one no consulting firm is saying out loud. The PE backers on the Anthropic side — Blackstone, Hellman & Friedman, Apollo, General Atlantic, Leonard Green, GIC, Sequoia — collectively own thousands of mid-market companies. The PE backers on the OpenAI side — TPG, Brookfield, Advent, Bain — own thousands more. Both JVs have a built-in customer pipeline before they have written a single line of customer code. The Big Three consultancies do not own their customers. They sell to them. When the lab and the asset manager are on the same cap table as the buyer, the consulting middle gets compressed first. The first $50 billion firm to fall in this transition is not a foundation model lab. It is a Big Four consulting house.

What to watch for the next 90 days. First, whether the OpenAI vehicle and the Anthropic vehicle stay distinct or one of them eats a Tier 2 consultancy outright. Second, whether Accenture, Deloitte, or PwC — the existing Claude Partner Network members — publicly respond, because Anthropic just put a competitor in the same channel they pay to be in. Third, whether the JVs publish customer logos. Fourth, whether the FDE headcount becomes public. Forward-deployed engineers are the unit economics of this whole bet. Fifth, whether a second consortium forms around Google or xAI. The market clearly supports two stacks. Three would be a phase change.

The takeaway is the line worth quoting. The frontier labs spent two years selling tokens. Today they decided that selling tokens is not enough. They are coming for the consulting layer above the tokens, and they brought the asset managers who own the customer base with them. McKinsey did not lose this round. McKinsey just stopped being the default.