OpenAI and Anthropic Are Coming for AI Services. Choose Wisely.

OpenAI and Anthropic Are Coming for AI Services. Choose Wisely.

OpenAI just acquired a services company. Anthropic built one from scratch with Blackstone and Goldman Sachs. The two biggest frontier model labs on the planet looked at the AI landscape and came to the same conclusion: models alone are not enough. You need people who can deploy them.

We have been saying this for two years. Virgent AI was the first AI services firm in the DMV built on this thesis. We always believed this day would come. And now it is here.

This is tremendous validation. It means the market we have been building in — the deployment layer, the implementation work, the messy business of turning AI into actual outcomes — is now a $5.5 billion-dollar land grab between the two biggest labs on Earth. That is exciting. It means we were right.

But let me be clear about something. We did not start Virgent AI to watch from the sidelines. We started this thing to get ahead, create value, and acquire the others in our wake. That vision has not changed.

Here is what actually happened, sourced and cited.

OpenAI Launches the Deployment Company, Acquires Tomoro

On May 11, 2026, OpenAI announced the OpenAI Deployment Company, a new enterprise AI deployment and services unit. As part of that launch, OpenAI agreed to acquire Tomoro, an applied AI consulting and engineering firm that brings roughly 150 Forward Deployed Engineers and Deployment Specialists into the fold.

The Deployment Company will be majority-owned and controlled by OpenAI, backed by more than $4 billion in initial investment from a 19-firm partnership led by TPG, with Advent, Bain Capital, and Brookfield as co-leads. Axios reported the new unit is valued at $14 billion. It is structured around embedding FDEs into organizations to redesign workflows and deploy AI systems.

Reuters independently confirmed the acquisition and added that Tomoro was formed in 2023 in alliance with OpenAI, with clients including Mattel, Red Bull, Tesco, and Virgin Atlantic. Bloomberg and Skift also covered the deal, noting Tomoro's work on Virgin Atlantic's AI concierge as a proof point.

Anthropic Builds a Parallel Services Company

A week earlier, on May 4, Anthropic announced a new AI services company formed with Blackstone, Hellman & Friedman, and Goldman Sachs. The model: Anthropic Applied AI engineers work alongside the firm's engineering team to identify high-impact use cases, build custom solutions, and support customers over time.

CNBC reported the venture is valued at $1.5 billion, including $300 million commitments each from Anthropic, Blackstone, and Hellman & Friedman. Fortune called it a direct shot at the consulting industry. The Blackstone press release confirms the structure — a standalone, AI-native enterprise services firm with additional backing from General Atlantic, Leonard Green, Apollo Global Management, GIC, and Sequoia Capital.

TechCrunch noted that both OpenAI and Anthropic announced enterprise services ventures within the same week — a clear signal the entire industry sees deployment as the next battleground.

The Pattern Is Unmistakable

Reuters reported on May 5 that both OpenAI and Anthropic had created private-equity-backed vehicles in talks to acquire AI services companies. The thesis is simple: enterprise AI adoption requires skilled, labor-intensive deployment work. Models do not deploy themselves.

This is not "AI labs doing consulting." It is a strategic land grab for the deployment layer — implementation, workflow transformation, integration, governance, and ongoing production support.

Source	What It Confirms
OpenAI official announcement	Deployment Company; Tomoro acquisition; 150 FDEs; $4B+ investment
Reuters, May 11	Independent confirmation of OpenAI unit and Tomoro deal
Axios	$14B valuation; PE-backed structure
Bloomberg	Tomoro acquisition details; Virgin Atlantic, Supercell clients
Anthropic announcement	New enterprise AI services company with Blackstone, H&F, Goldman
CNBC	$1.5B valuation; $300M commitments from each founding partner
Fortune	Positioned as direct challenge to consulting industry
Blackstone press release	Standalone structure, investor consortium, enterprise positioning
TechCrunch	Both labs launching services ventures in the same week
Reuters, May 5	Both companies pursuing AI services acquisitions

Hats Off. Now Let Me Tell You the Problem.

I want to say this plainly: hats off to OpenAI. What they are building is impressive. The scale, the ambition, the speed of execution. Same for Anthropic. These are serious moves by serious organizations and it validates everything we have been grinding on.

But here is the thing. When OpenAI deploys engineers into your organization, they are deploying OpenAI. When Anthropic embeds Applied AI engineers, they are embedding Claude. That is not a services engagement. That is vendor lock-in wearing a consulting badge.

You will get one mono stack. Their models. Their AI dev tooling. Their orchestration. Their governance frameworks. Everything dogfooded from a single ecosystem. And you will end up trapped in a monoculture — not because it was the best fit for your problems, but because it was the only option on the table.

Forrester warned about this exact pattern in August 2025: "Vendors are leveraging their entrenched positions to end discounting and push high-margin AI SKUs." Salesforce's Marc Benioff openly told investors that charging $2 per AI agent conversation would be "a very high margin opportunity." That is the playbook. Get in. Get entrenched. Extract.

Your business is unique. Your data is unique. Your workflows, your compliance requirements, your competitive advantages — all unique. So why would you hand all of that to a company whose primary incentive is to maximize consumption of their own API?

We have seen this movie before. It was called Salesforce. An $18,000/year subscription that actually costs $47,000 in year one once you add implementation, data migration, and required add-ons — with 7% annual price increases baked into the contract. The companies that went all-in spent the next decade paying consultants to untangle the mess.

Why Virgent AI Is the Better Bet

We are sovereign. We are not owned by a model lab. We are not backed by a PE consortium whose portfolio companies need to consume more tokens. We answer to our customers and our outcomes. That is it.

And because we are sovereign, we will always champion a vendor agnostic approach. Full stop. We pick the best models, the best tooling, and the best architecture for your specific problem — to maximize value and outcomes for you, not to shoehorn you into a walled garden. Sometimes that is OpenAI. Sometimes that is Anthropic. Sometimes it is an open-source model running on your own infrastructure. The answer depends on your requirements, not our partnerships.

We have shipped this work. Not decks. Not proofs of concept that die in staging. Production systems with measurable outcomes.

• 60 Days: Kickoff to ROI. Production AI agent deployed in 2 weeks, replacing a failing chatbot. $10,000+ in monthly savings. 50% ticket reduction. $120K+ in annualized savings. ROI in under 60 days.
• Agentic Layers Accelerating Sales and Hiring. Multi-model agentic workflows embedded in websites and sales funnels that qualify leads, screen candidates, triage inquiries, and route the right people to each other — automatically and intelligently.
• The Agentic Stack. One engineer. Twelve production systems. Zero incidents. We did not just write about the agentic stack — we shipped every layer of it.
• Multi-Agent AI Orchestration. Agent systems that collaborate, vote, and evolve. Live demos you can try right now.

That is not marketing copy. Those are case studies you can read with specific numbers, specific architectures, and specific outcomes.

What We Bring That They Cannot

• Product management excellence. We do not just build AI features. We identify the use cases that actually move your business metrics, scope them properly, and ship them with measurable outcomes. I taught Product Management and AI at an art college. I have led design teams and architected $100M+ modernizations. That discipline is in everything we ship.
• AI expertise across the full stack. From model selection and fine-tuning to agent orchestration, MCP and A2A integration, and production deployment. We adopted MCP and A2A before most teams had names for what they were trying to build. We built Cadderly as a coordination agent incorporating both protocols alongside custom intent recognition and multi-model orchestration.
• Best-in-class delivery. 100+ commercial and government projects shipped through AltonTech. Real outcomes. Real ROI. Not slide decks.
• Local presence. We are in Annapolis. We are in the DMV. We show up. We are not a Forward Deployed Engineer parachuting in from San Francisco for six weeks.
• Vendor agnosticism that protects your future. We will never lock you into a single model provider. Your architecture stays portable. Your leverage stays intact. When the next model breakthrough happens — and it will — you can adopt it immediately instead of renegotiating a contract.
• Transparent pricing. $450/hour. $4,500 for a one-day workshop. $6,000/month retainer. First call is always free. No hidden fees. No enterprise subscription traps.

The Big Picture

The frontier labs entering services is the strongest possible validation that this work matters. That the deployment layer is where the real value gets created. That models without implementation are just expensive autocomplete.

$5.5 billion in combined investment says the market we have been building in is real. We are genuinely excited to see OpenAI and Anthropic confirm what we have known since day one. It raises the profile of every company doing this work well. It brings more attention, more budget, and more urgency to the deployment layer.

We started Virgent AI to get ahead, create value, and acquire the others in our wake. That vision has not changed. These moves by the labs do not scare us. They sharpen us. And while we respect what they are building, we are going to keep doing what we do best — championing a vendor agnostic approach that maximizes value and outcomes for our customers, rather than funneling them into someone else's walled garden.

The choice is clear for every organization adopting AI: do you want a partner whose incentive is to sell you more of their own product? Or do you want a partner whose only incentive is to ship the best possible outcome for your business?

We were here first. But being first is a whack value proposition. Being the best is the value prop you want to grind for. We will be here after the hype cycle. And we will keep doing what we have always done — shipping outcomes for real organizations with real problems.

If you want a partner who works for you and not for a model provider's growth targets, let's talk.