In the first two weeks of May 2026, OpenAI and Anthropic each launched billion-dollar AI consulting ventures designed to embed engineers directly inside companies and deploy AI into production workflows. Combined, these two ventures represent $5.5 billion in committed capital and signal that the AI industry has recognized a fundamental truth: selling software is no longer enough. Companies need people who can put it to work.
The problem for mid-market companies, those with 100 to 1,000 employees, is that neither venture was built for them.
OpenAI’s Deployment Company launched on May 11, 2026 with $4 billion in backing and a $10 billion pre-money valuation (Axios, 2026). Anthropic’s unnamed venture, announced May 4, raised $1.5 billion from Blackstone, Goldman Sachs, and Hellman & Friedman (CNBC, 2026). Both follow the same model pioneered by Palantir: embed Forward Deployed Engineers inside client organizations to build AI systems against real data in real workflows.
This is meaningful news. It validates an approach to AI deployment that AiBT Advisory was built around. And it clarifies what the market looks like for every mid-market leader currently evaluating where AI fits their company.
What did OpenAI and Anthropic actually launch?
OpenAI created a standalone business unit called the OpenAI Deployment Company, or DeployCo. It operates as a separate entity, majority-owned by OpenAI, backed by 19 investment firms, consultancies, and system integrators. The partnership is led by TPG, with Advent, Bain Capital, and Brookfield as co-leads. McKinsey, Bain & Company, and Capgemini are also investors.
To staff the operation from day one, OpenAI acquired Tomoro, an applied AI consulting firm, bringing approximately 150 Forward Deployed Engineers and deployment specialists into the new company (OpenAI, 2026). DeployCo’s stated mission is to help organizations “identify where AI can make the biggest impact, redesign organizational infrastructure and critical workflows around it, and turn those gains into durable systems.”
Anthropic took a different approach. Its venture, which has not been publicly named, is a standalone entity formed in partnership with Blackstone, Hellman & Friedman, and Goldman Sachs. Anthropic, Blackstone, and Hellman & Friedman each contributed roughly $300 million, with Goldman Sachs contributing $150 million (Fortune, 2026). The venture is also backed by Apollo Global Management, General Atlantic, GIC, Leonard Green, and Sequoia Capital.
The Anthropic venture will deploy Claude, Anthropic’s AI model, directly inside businesses, starting with companies owned by the investment firms in the consortium. Goldman Sachs’s Marc Nachmann described the goal as helping “mid-market companies deploy Anthropic’s AI solutions to drive meaningful impact in their business” (GIC Newsroom, 2026).
Both ventures follow what the industry calls the Forward Deployed Engineer model.
What is the Forward Deployed Engineer model?
The Forward Deployed Engineer (FDE) model was created by Palantir Technologies in the early 2010s. Instead of selling software and letting the customer figure out how to implement it, Palantir embedded its own engineers directly inside client organizations to build production systems against real data in real operational environments.
As Palantir described it, “FDEs work in small teams and own end-to-end execution of high-stakes projects” (Palantir, 2022). These engineers would build working prototypes within 30 days, expand into production within 90 days, and generate internal champions who would advocate for contract expansion. Until 2016, Palantir had more FDEs than traditional software engineers (Pragmatic Engineer, 2025).
The model worked. Palantir IPO’d at $19 per share in 2020, dropped to $6 by 2022, and then delivered approximately 640% returns over five years. Analysts initially questioned whether the embedded engineering approach could scale. The revenue numbers settled the debate.
The reason FDE-deployed AI systems are so sticky is that every deployment is custom. The integrations are built around the client’s specific data, workflows, and compliance requirements. Switching to a different vendor means rebuilding everything. This is the business model both OpenAI and Anthropic are now pursuing at scale: high cost to acquire, very high retention, very high contract value.
Why were these ventures launched now?
The timing reflects a specific market reality: AI adoption is widespread, but AI deployment into production remains rare.
Eighty-eight percent of organizations now use AI in at least one business function (McKinsey, 2025). But more than 80% of those organizations report no tangible EBIT impact from their AI initiatives (AmplifAI, 2026). The MIT NANDA initiative’s 2025 “GenAI Divide” report found that while 60% of organizations had evaluated enterprise-grade AI tools, only 20% reached the pilot stage, and just 5% made it to production.
The pattern is consistent across company sizes and industries: teams learn the tools, run a pilot, see the potential, and then have no infrastructure to put any of it into production. Forty-two percent of companies abandoned most of their AI initiatives in 2025, up from 17% in 2024 (Fullview, 2025). Only 26% of organizations report having the capabilities to move beyond proof-of-concept to production deployment.
OpenAI and Anthropic saw this gap and recognized the revenue opportunity. As Futurum Group analyst David Nicholson put it, the companies understand that “the real revenue opportunity exists where the rubber meets the road,” in the people who can “help enterprises take these fancy toys and turn them into business value” (AI Business, 2026).
Both ventures are backed by private equity firms with captive distribution into hundreds of portfolio companies. Anthropic’s consortium alone spans Blackstone, Goldman Sachs, Apollo, General Atlantic, and Sequoia, each managing portfolios with dozens to hundreds of companies that need AI deployment.
Are these ventures built for mid-market companies?
Anthropic’s venture explicitly mentions mid-sized companies as a target. Goldman Sachs’s Nachmann described the venture as a way to “democratize access to forward-deployed engineers for companies that currently cannot afford either the talent or the fees charged by large consulting firms” (Fortune, 2026).
OpenAI’s DeployCo is broader in its framing, describing its target as “organizations working on complex problems in demanding environments” (OpenAI, 2026). The partners backing DeployCo, including TPG, Bain Capital, and Brookfield, collectively sponsor more than 2,000 businesses.
In practice, the economics of the FDE model create a natural floor. Forward Deployed Engineers at US AI-native companies earn $180,000 to $280,000 in total compensation (Florian Nègre, 2025). A typical FDE engagement embeds a small team inside the client for three to six months. At big consulting firms, AI strategy-plus-prototype programs start at $1 million and can run to $30 million (Aiken House, 2026). Even smaller-scale AI consulting engagements at firms like Accenture, Deloitte, and IBM start at $500,000.
The investment structures reinforce this. OpenAI’s investors are guaranteed a minimum 17.5% annual return over five years (Axios, 2026). Anthropic’s backers include firms managing trillions in assets who expect the venture to be “a compelling investment opportunity” (Goldman Sachs, 2026). These returns require high-value enterprise engagements, not $50,000 mid-market projects.
The World Economic Forum described mid-market companies as “the missing middle” in AI adoption. The demand is there. The supply chain is not.
The World Economic Forum described mid-market companies as “the missing middle” in AI adoption: large enough to have real operational complexity where AI creates measurable value, but too small to access the consulting infrastructure built for Fortune 500 organizations (WEF, 2026). RSM’s 2025 Middle Market AI Survey found that 91% of middle market firms now use generative AI, but 70% recognize the need for external support to maximize the value of their AI investments.
The demand is there. The supply chain is not.
What does this mean for companies with 100 to 1,000 employees?
These announcements validate the deployment model. They also make clear who it was sized for.
Both ventures are backed by private equity firms expecting returns on billions in committed capital, and the consulting firms in their orbit price their work accordingly. The embedded deployment approach works. But for a company with 300 or 500 employees, these are not accessible options. That leaves a real gap in the market: companies large enough to have the operational complexity where AI creates measurable value, but too small to show up on the radar of a $10 billion consulting vehicle.
AiBT Advisory was built specifically for companies with 100 to 1,000 employees, applying that same embedded approach at a scale where we work directly with your leadership team, your operators, and your data. Our engagement starts with an AI Opportunity Audit where we talk to your leadership team, your department heads, and the people doing the day-to-day work. We look at your tech stack, your data, and the workflows where time and money disappear into manual effort. Then we deploy AI into the workflows that create the most measurable impact and stay through adoption and ongoing operations.
Both ventures are also tied to their own AI models. For mid-market companies, being locked to a single provider creates real risk while the landscape is still shifting this fast. AiBT Advisory is model-agnostic. We deploy whatever tools and models work best for each use case. When something better comes along, we swap it in. The systems we build are designed to keep working and improving as the technology changes, without tying your company to any one provider.
Most companies that have tried AI internally ran into the same wall: the pilot worked, the team saw the potential, and then nobody built the bridge to production. There was no ownership structure, no way to track whether people were actually using the tools, no integration with existing systems, and no one accountable for making it work inside the workflows your teams run every day. AiBT Advisory was built to close that gap. We configure, integrate, and deploy AI into your workflows, connected to the tools your teams already use. We train your people, build the documentation, and track adoption at 30, 60, and 90 days so you can see what is working and where it needs attention.
What should mid-market leaders do next?
If your company has 100 to 1,000 employees, these announcements should accelerate your timeline, not delay it. Waiting for billion-dollar ventures to work their way down to your scale could take years. The companies that move now build the advantage.
Every AiBT Advisory engagement starts with the AI Opportunity Audit. We spend two to three weeks inside your company, talking to the people who actually run it. You walk away with a prioritized AI Opportunity Map showing where AI creates the most measurable value across your workflows, with ROI projections for each one. A clear picture of what is worth building and what it will produce.
From there, we handle the full engagement: selecting the highest-impact workflows, building and deploying the systems into production, training your teams, and staying on retainer to keep everything working as your company evolves.
The biggest AI companies just made a $5.5 billion bet that deployment is the business. If your company has between 100 and 1,000 employees and you’re evaluating where AI fits, we should talk.