TL;DR
OpenAI and Anthropic made parallel moves in early May 2026 to push deeper into enterprise AI services, according to Thorsten Meyer AI. The reported shift centers on forward-deployed engineers who embed with companies to move AI from pilots into production.
OpenAI and Anthropic moved within roughly 72 hours in early May 2026 to build enterprise AI deployment operations modeled on Palantir’s forward-deployed-engineer approach, according to Thorsten Meyer AI, a shift that could change how frontier AI labs sell to companies and compete with consultants.
Thorsten Meyer AI reported that Anthropic announced a $1.5 billion enterprise-services venture with Blackstone, Hellman & Friedman, and Goldman Sachs to embed Claude inside mid-market companies. Hours later, according to the same source material, OpenAI announced a $4 billion Deployment Company, called DeployCo, at a $10 billion pre-money valuation with 19 investment partners.
The OpenAI structure also included the immediate acquisition of consulting firm Tomoro, bringing 150 forward-deployed engineers into the company on day one, according to the report. The model described in the source material places engineers inside client companies to study workflows, build production software around frontier models, and remain involved until the deployment works.
The central argument in the report is that AI labs are moving into the services layer because enterprise AI adoption is being slowed less by model quality than by integration, security review, evaluation systems, and process redesign. The report cites MIT research saying 95% of generative AI pilots fail to move beyond the experimental phase.
Why It Matters
The shift matters because it suggests leading AI labs are no longer treating enterprise adoption as a software licensing problem alone. If companies struggle to put AI into production, the firms that control both the model and the deployment process may capture more of the value created by AI systems.
The report frames the economics around a software-to-services ratio: for every dollar companies spend on software, they spend roughly six on services. If accurate, that makes implementation, integration, workflow redesign, and change management a larger prize than model access by itself.
For customers, the move could bring faster production deployments but also tighter vendor dependency. Systems built around a lab’s model, tools, evaluation process, and embedded engineering team may be harder to replace later.
Autonomous AI-Driven Enterprise Software From Development to Deployment
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background
Palantir’s forward-deployed-engineer model grew out of work with defense, intelligence, and large institutional customers. In that model, engineers work close to the customer’s operations and build software around specific workflows rather than simply handing over a platform.
Thorsten Meyer AI says OpenAI and Anthropic are applying that approach to the broader enterprise market. The report argues that this is part of a wider change in the AI business: model access is becoming less differentiated, while deployment capability, customer integration, and recurring usage may become more important sources of revenue.
“the model isn’t the bottleneck, deployment is”
— Thorsten Meyer AI
“almost line for line”
— Thorsten Meyer AI, describing the Palantir comparison
“resembles consulting more than pure software licensing”
— Thorsten Meyer AI
AI at the Edge: Solving Real-World Problems with Embedded Machine Learning
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What Remains Unclear
Several details remain unclear from the provided source material, including the exact ownership structures of the reported ventures, the contractual terms with investment partners, how revenue will be shared, and whether OpenAI or Anthropic have independently detailed the long-term margin profile of these deployment businesses.
The largest open question is whether forward-deployed engineering becomes a repeatable product engine or remains a labor-heavy services business. If each customer requires extensive bespoke work, margins could look more like consulting than software.
AI for Nurses: The Practical Guide to HIPAA-Compliant AI Tools, Documentation Workflows, and Ethical Integration for Registered Nurses and Nurse Practitioners (AI for Professionals)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What’s Next
The next test is execution: whether OpenAI and Anthropic can turn embedded engineering work into repeatable products, higher enterprise usage, and durable customer retention. Investors and customers will be watching deployment speed, production adoption rates, margins, and whether customers expand usage after the first implementation.
Your AI Survival Guide: Scraped Knees, Bruised Elbows, and Lessons Learned from Real-World AI Deployments
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the actual news development?
OpenAI and Anthropic reportedly made parallel moves in early May 2026 to create enterprise deployment operations built around embedded engineers who help customers put AI systems into production.
Why are AI labs moving into services?
The report says enterprise adoption is being slowed by integration, security review, evaluation, and workflow redesign rather than model performance alone. Services are also described as a much larger spending category than software.
What is a forward-deployed engineer?
In the model described by Thorsten Meyer AI, a forward-deployed engineer works inside or close to a client organization, learns its workflows, builds software around real operational problems, and stays involved until the system works in production.
What is the main risk?
The risk is that the model may require large amounts of human engineering time for each customer. That could limit scale and reduce margins if the work does not become repeatable across clients.
Why does Palantir matter here?
Palantir is cited because its business has long used embedded technical teams to build and deploy software around customer operations. The report says AI labs are now applying a similar structure to enterprise AI.
Source: Thorsten Meyer AI
