← Back to feed

Anthropic is close to finalising a $1.5 billion joint venture with three Wall Street firms: Blackstone, Hellman & Friedman, and Goldman Sachs. The Wall Street Journal and Reuters reported the deal structure. Fortune confirmed Anthropic framed the new entity as a direct challenge to the consulting industry.

Structure

• Anthropic: ~$300M
• Blackstone: ~$300M
• Hellman & Friedman: ~$300M
• Goldman Sachs: ~$150M
• Total: ~$1.5B

The new company is intended to function as an AI deployment arm, not a reseller. It will not hand customers a Claude API key — it will go inside companies and rebuild customer support, finance, legal review, coding, and internal research workflows around LLMs. Private equity firms own large portfolios of operating companies, have tight cost discipline, and can force software changes faster than public companies with procurement bureaucracy.

The Strategic Logic

Anthropic gets distribution it cannot build organically. Signing up individual enterprise customers one at a time is slow. PE firms like Blackstone and H&F control dozens or hundreds of portfolio companies simultaneously; a single JV relationship becomes a multi-company deployment channel.

Wall Street gets a stake in the services layer sitting between the model and the enterprise. Claude API pricing is transparent and declining; the margin is in implementation and workflow redesign.

The target customer — a PE-backed company being pushed by its sponsor to cut headcount and automate — is also the customer most likely to deploy AI aggressively. That makes the portfolio a large-scale test bed for what Claude-native workflows actually look like at scale.

The Broader Pattern

This is the second major deployment-arm structure Anthropic has backed. OpenAI launched a comparable vehicle on the same day as an earlier Anthropic announcement, with capital from different PE sources. The two moves together signal that frontier labs have concluded they cannot capture enterprise value by selling inference alone.

The model-to-business-process gap is where the money is. Consulting firms — Accenture, McKinsey, Deloitte — have known this for decades and are now the primary competition for AI deployment revenue. Anthropic’s JV positions it to attack that market directly, with its own capital at risk, rather than relying on third-party system integrators who might route work toward cheaper models.

For Anthropic, the deeper value may be the data. A deployment arm embedded in PE portfolio companies generates workflow-level signal that an API relationship never does — which tasks are actually being automated, where models fail, what prompts work. That feedback loop is hard to replicate from API logs alone.