Anthropic just crossed a critical threshold in enterprise AI adoption: 73% of companies buying AI tools for the first time are now choosing Anthropic over OpenAI, according to billing data from Ramp tracking 50,000+ businesses. That's up from a 50/50 split just four months ago in January 2025.

The shift isn't accidental. On May 4, 2026, Anthropic announced a $1.5 billion joint venture with three of Wall Street's most powerful investors—Blackstone, Hellman & Friedman, and Goldman Sachs—to build what insiders are calling the "McKinsey of AI." The unnamed enterprise services company will embed engineers directly into mid-sized businesses to deploy Claude AI across their core operations, starting with 275 Blackstone portfolio companies.

For CFOs evaluating AI budgets and CTOs selecting vendors, this deal reveals a fundamental shift in how enterprise AI gets implemented—and why first-time buyers are betting on Anthropic instead of OpenAI.

The Numbers Tell the Story

The market share reversal happened fast:

• January 2025: Anthropic and OpenAI split first-time enterprise buyers 50/50
• March 2026: Anthropic captures 73%+ of new buyer spending
• May 2026: Wall Street commits $1.5 billion to scale Anthropic's enterprise model

Ramp, a corporate card and billing platform serving 50,000+ companies, tracks AI spending patterns across industries. The data shows Anthropic isn't just winning new customers—it's fundamentally reshaping enterprise procurement behavior.

The $1.5 billion joint venture breakdown:

• Anthropic: $300 million
• Blackstone: $300 million
• Hellman & Friedman: $300 million
• Goldman Sachs Asset Management: $150 million
• Additional backers: Apollo, General Atlantic, Leonard Green, GIC, Sequoia Capital ($450 million)

"Enterprise demand for Claude is significantly outpacing any single delivery model," said Krishna Rao, Anthropic's Chief Financial Officer, in the May 4 announcement.

Why First-Time Buyers Are Choosing Anthropic

The embedded engineering model solves the implementation bottleneck that kills most AI pilots.

Traditional AI vendors sell software licenses. Systems integrators like Accenture or Deloitte charge $300-500/hour for consultants to build on top of those licenses. The new Anthropic model collapses that gap: engineers work directly inside businesses to redesign workflows and integrate Claude into core processes.

"Having the model alone doesn't change your workflows or how you operate," Marc Nachmann, Goldman's global head of asset and wealth management, told CNBC. "You need people who can combine the technology with what's actually happening in the business and implement those changes."

Consider a real-world example from Anthropic's announcement: A multi-site healthcare network where physicians spend hours daily on documentation, medical coding, prior authorizations, and compliance reviews. The engagement starts with Anthropic engineers and the company's technical team sitting down with clinicians and IT staff to build tools that fit existing workflows.

The clinicians know where time disappears. The engineers build around that knowledge. The result: more time for patient care, less time on administrative overhead.

This is fundamentally different from how OpenAI sells enterprise AI.

OpenAI's enterprise strategy centers on ChatGPT Enterprise subscriptions at $60/user/month and API access for developers. Anthropic's new model embeds engineers on-site to transform operations—closer to McKinsey's transformation consulting than SaaS licensing.

The $9 Trillion Efficiency Play

Wall Street isn't betting $1.5 billion on software licenses. They're betting on operational leverage across their portfolio companies.

Jon Gray, Blackstone's president, explained the math on CNBC: "Labor is a $60 trillion annual cost. If you can make people 15% more efficient, that's $9 trillion."

Blackstone owns 275 companies across healthcare, manufacturing, financial services, retail, and real estate. Hellman & Friedman and Goldman Sachs manage thousands more. The joint venture gives them first access to Anthropic's latest models before public release, embedded engineering expertise, and a repeatable AI transformation playbook.

The scale of private equity's AI push is staggering:

• Blackstone: $150 billion in data centers globally, $160 billion pipeline
• Blackstone portfolio companies: LLM spending up 15x year-over-year
• Private equity holding periods: 40% of companies held 5+ years (up from 29% in 2019)

"The demand is off the charts," said Brian Mulberry, chief market strategist at Zacks Investment Management. "When computing power grows and more data centers come online, the use of AI as a tool to drive productivity becomes very scalable."

Private equity firms need to generate returns on aging assets. AI-driven efficiency gains offer a path to justify valuations and accelerate exits. But they need implementation expertise, not just software licenses.

What This Means for CFOs and CTOs

If you're evaluating AI vendors in 2026, the Anthropic-OpenAI choice isn't just about model performance anymore. It's about implementation models.

For CFOs: The Total Cost Question

Anthropic's embedded engineering model shifts AI spending from CapEx to transformation services.

Traditional enterprise AI deals stack costs:

• Software licenses: $60/user/month (ChatGPT Enterprise) or per-token API pricing
• Systems integrator consulting: $300-500/hour for Accenture, Deloitte, PwC
• Internal IT resources: 2-4 FTEs for integration and maintenance
• Training and change management: $50-150/employee

Anthropic's joint venture bundles the engineering layer directly. For mid-sized companies (500-5,000 employees), this could reduce total implementation costs by 30-40% by eliminating the systems integrator markup and accelerating time-to-value.

The ROI question shifts from "How much per token?" to "How much efficiency gain per dollar invested?"

If Anthropic can deliver 10-15% productivity improvements across 1,000-employee organizations, the payback period drops to 6-12 months versus 18-36 months for traditional AI pilots.

For CTOs: The Implementation Risk Question

78% of enterprise AI pilots die before production, according to industry benchmarks. Anthropic's embedded model directly targets that failure rate.

The typical AI pilot failure pattern:

1. Vendor demo looks promising (1-2 weeks)
2. IT team spins up API access (1-2 weeks)
3. Engineers build proof-of-concept (4-8 weeks)
4. Business units don't adopt the tool (workflows don't match)
5. Pilot expires, budget reallocates

Anthropic's engineers work alongside your team from day one. They understand where Claude can have the biggest impact, build custom solutions, and support over the long term. This isn't a proof-of-concept—it's operational transformation with committed resources.

For technical leaders, the question becomes: Do you want a software license and a consulting contract, or do you want embedded engineers who share accountability for business outcomes?

The Competitive Landscape

OpenAI still dominates consumer AI (ChatGPT) and developer tools (API). But Anthropic is winning the enterprise transformation market.

The Ramp data shows first-time buyers shifting to Anthropic at 3x the rate of OpenAI. The reasons stack up:

1. Safety and compliance focus: Anthropic's Constitutional AI framework resonates with regulated industries (healthcare, financial services)
2. Claude Partner Network: Existing partnerships with Accenture, Deloitte, PwC for Fortune 500 deployments
3. Enterprise-first model design: Claude is built for long-context enterprise workflows (200K tokens vs GPT-4's 128K)
4. Embedded engineering: The new joint venture accelerates mid-market adoption where OpenAI has fewer resources

OpenAI's $110 billion valuation (April 2026 funding round) reflects its leadership in consumer AI and API-first developer tools. Anthropic's $1.5 billion enterprise services bet reflects where Wall Street sees the next $9 trillion in value creation: operational efficiency in real-world businesses.

Decision Framework for Enterprise Buyers

If you're choosing between Anthropic and OpenAI in 2026, ask these questions:

1. What's your implementation model?

• Self-service API integration? OpenAI's developer-first tools may be faster
• Enterprise transformation with embedded support? Anthropic's joint venture model fits better

2. What's your industry regulatory profile?

• Highly regulated (healthcare, finance, government)? Anthropic's Constitutional AI and safety focus reduce compliance risk
• Lower regulatory burden? OpenAI's API flexibility may offer more customization options

3. What's your company size and technical maturity?

• Large enterprise (10,000+ employees)? Claude Partner Network (Accenture, Deloitte) handles scale
• Mid-sized company (500-5,000 employees)? Anthropic's new embedded engineering model targets this segment
• Small business (<500 employees)? OpenAI's ChatGPT Enterprise at $60/user/month is simpler to deploy

4. What's your timeline and risk tolerance?

• Need production results in 3-6 months? Embedded engineers reduce pilot-to-production failure rates
• Can afford 12-18 month pilots? OpenAI's API-first model gives more architectural control

The Ramp data suggests first-time buyers are prioritizing implementation support over architectural flexibility. That's a rational choice when 78% of AI pilots fail—better to pay for embedded expertise than risk a failed proof-of-concept.

What Happens Next

The unnamed Anthropic-Blackstone-H&F-Goldman Sachs joint venture will launch in Q3 2026, starting with private equity portfolio companies before targeting the broader mid-market.

Expect to see:

1. Playbook development across 5-10 initial customers (healthcare, manufacturing, financial services)
2. Case studies demonstrating 10-15% productivity gains and 6-12 month ROI
3. Scaling beyond PE portfolios to any mid-sized company willing to pay for embedded AI transformation
4. OpenAI's response: Likely partnerships with systems integrators or their own embedded services model

For enterprise buyers, the market is bifurcating: consumer AI stays API-first and developer-centric (OpenAI's strength), while enterprise transformation moves toward embedded engineering and operational accountability (Anthropic's bet).

The 73% market share shift among first-time buyers isn't a fluke. It's a signal that enterprise AI is moving from "buy software" to "buy outcomes." Wall Street just bet $1.5 billion that Anthropic can deliver those outcomes faster and more reliably than anyone else.

Sources

1. Anthropic Partners with Blackstone, Hellman & Friedman, and Goldman Sachs to Launch Enterprise AI Services Firm - Anthropic Official Announcement, May 4, 2026
2. Anthropic teams with Goldman, Blackstone and others on $1.5 billion AI venture targeting PE-owned firms - CNBC, May 4, 2026
3. Wall Street's $1.5 billion plan to build the 'McKinsey of AI' - Business Insider, May 4, 2026
4. Ramp data: Anthropic is now capturing 73%+ of all spending among companies buying AI tools for the first time - Techmeme/Axios, March 18, 2026