In a coincidence so perfectly timed it almost feels staged, Anthropic and OpenAI both launched joint ventures for enterprise AI services on Monday — within hours of each other. If the AI industry was a movie, this would be the scene where the two rival gangs show up to the same bank heist.
The Deals
Anthropic announced a joint venture valued at $1.5 billion, with founding partners Blackstone, Hellman & Friedman, and Goldman Sachs. Each of those three is putting in $300 million. The backer roster reads like a Wall Street who’s who: Apollo Global Management, General Atlantic, GIC, Leonard Green, and Sequoia Capital.
Meanwhile, OpenAI’s venture — charmingly named “The Development Company” — is operating at a different scale entirely: $10 billion valuation, $4 billion raised from 19 investors including TPG, Brookfield Asset Management, Advent, and Bain Capital. Not a single investor overlaps between the two deals.
Same day. Same playbook. Ten times the price tag.
Why This Matters
Both ventures follow the same logic: AI labs partner with private equity and asset managers to sell enterprise AI deployments into their portfolio companies. The investors get preferred access and a cut of the contracts. The AI labs get a dedicated sales channel with built-in customers. Everyone wins — except maybe the companies that end up paying enterprise prices for AI tools that may or may not deliver.
The most interesting detail? Both ventures are embracing the forward-deployed engineer (FDE) model — the same approach Palantir used to build its empire. As Anthropic put it, engagements begin with engineers “sitting down with clinicians and IT staff to build tools that fit into the workflows that staff already use.”
Translation: the AI labs have realised that selling API access isn’t enough. They need bodies on the ground, customising deployments for each enterprise. The “just call our API” era is officially over.
The Bigger Picture
This is the AI industry growing up — or at least dressing like it. Both companies are shifting from “build the smartest model” to “build the smartest sales operation.” The joint venture structure lets them do this without diluting their core research focus, while giving Wall Street a piece of the action.
But here’s the question nobody’s asking: if enterprise AI requires this much hand-holding, how good is the product really? The FDE model worked for Palantir because their product was genuinely complex and bespoke. If Anthropic and OpenAI need forward-deployed engineers to make their AI useful in hospitals and banks, that says something uncomfortable about the gap between demo and reality.
NZ Angle
For New Zealand businesses, this is actually good news — sort of. The enterprise JV model means these companies will eventually want customers beyond the Fortune 500. Mid-market companies in Australasia could benefit from the same FDE approach, though they’ll be waiting in line behind the Blackstone portfolio.
The less good news: when the world’s two leading AI labs are structuring billion-dollar deals around “helping enterprises deploy AI,” it’s a signal that deployment — not development — is now the bottleneck. If you’re a NZ company still “exploring” AI adoption, you’re not behind the curve. The curve is still being built.
🔍 THE BOTTOM LINE
Anthropic and OpenAI launching matching joint ventures on the same day isn’t just a PR moment — it’s a milestone. The AI industry has officially entered its corporate era. The question is whether this is the beginning of real enterprise AI adoption, or just the beginning of a very expensive sales exercise. Either way, Wall Street’s all-in.
