What changed
Between May 4 and May 18, four separate announcements from OpenAI, Anthropic, Meta and Dell Technologies converged on the same problem: enterprises are buying AI models but struggling to put them to work. Each company responded by building or expanding a unit whose job is to send engineers into client offices and rebuild workflows around the technology, rather than selling the technology and walking away.
On May 4, Anthropic announced a new enterprise AI services firm backed by Blackstone, Hellman & Friedman and Goldman Sachs, valued at $1.5 billion, with Anthropic, Blackstone and Hellman & Friedman each committing $300 million. Apollo Global Management, General Atlantic, Leonard Green, GIC and Sequoia Capital joined as additional backers. The firm will embed Anthropic engineers inside mid-sized companies, the segment Anthropic says lacks the in-house staff to deploy frontier models on its own.
A week later, on May 11, OpenAI launched the OpenAI Deployment Company, a majority-owned subsidiary capitalized with more than $4 billion and valued at $10 billion, according to OpenAI's own announcement. The new unit acquired Tomoro, a London-based applied AI consulting firm founded in 2023 in partnership with OpenAI, bringing roughly 150 forward-deployed engineers whose client list includes Fidelity International, Virgin Atlantic, Tesco and the NBA. TPG leads a group of 19 investment firms, consultancies and system integrators backing the venture, with Advent, Bain Capital and Brookfield as co-lead partners.
Meta followed with plans for an Enterprise Solutions unit, structured around product managers who run client engagements, data engineers who prepare customer data, and software engineers who embed Meta's AI tools into client systems. Meta CTO Andrew Bosworth has called 2026 "a critical year" for the company's push to monetize its AI investment.
Then, on May 18, Dell Technologies World in Las Vegas served as the enterprise-adoption bookend. Dell said it added 1,000 new customers to its AI Factory program with NVIDIA in the prior quarter, bringing the total to 5,000, with Eli Lilly, Honeywell and Samsung named among the companies building AI infrastructure on the platform. Dell also introduced PowerRack, a turnkey rack-scale system combining compute, networking and storage that the company says can be operational within 6.5 hours of arrival, with the networking configuration due in September 2026 and the storage configuration in the second half of the year.
The common thread
None of these four moves is a model release. Each is a bet that the constraint on enterprise AI has shifted away from model capability and toward installation. Reporting has pointed to OpenAI, Anthropic and Microsoft collectively committing roughly $8 billion this year to deployment ventures of this kind, citing survey data in which 71% of executives at companies with at least $1 billion in annual revenue point to organizational readiness, not the technology itself, as the primary barrier to AI performance, versus 11% who cite the technology.
That statistic explains why three model labs that compete fiercely on benchmarks are now competing on the same non-model metric: how fast they can get an agent workflow running inside a specific company's specific systems. The forward-deployed engineer model, historically associated with Palantir, has become the template. Google, NVIDIA and now Meta have adopted variants of it alongside OpenAI and Anthropic.
Who this affects
For enterprise buyers, the practical effect is a narrowing of the gap between piloting a model and running it in production, provided they can afford the engagement. Anthropic's venture is explicitly aimed at mid-sized companies priced out of building their own AI engineering teams; OpenAI's Deployment Company inherits Tomoro's existing roster of large accounts. Both models depend on private equity and asset management partners, which suggests the near-term customer base skews toward portfolio companies of firms like Blackstone, TPG and Apollo rather than open market sales.
For the consulting and systems-integration industry, the acquisitions are a direct incursion. Tomoro was an independent applied AI consultancy before OpenAI bought it outright. Traditional integrators now compete with the model vendors themselves for the same deployment contracts.
For infrastructure vendors, Dell's numbers show the buildout continuing in parallel: a fivefold customer increase in the AI Factory program over the past year is an indicator of infrastructure demand distinct from, but complementary to, the deployment-services push. Enterprises adopting these forward-deployed engineering arrangements will still need somewhere to run the resulting workloads, and Dell's timing at its own conference on May 18 positioned the company to capture that side of the same spend.
Sources: Anthropic: Building a new enterprise AI services company, May 4, 2026 · CNBC · TechCrunch · OpenAI: OpenAI Deployment Company, May 11, 2026 · Bloomberg · AIwire/HPCwire · PYMNTS: Meta enterprise AI · PYMNTS: AI deployment spend · Dell Technologies World 2026 announcements · DataCenterDynamics: Dell PowerRack