OpenAI and Anthropic are running the same play at the same time, and it tells you something about where the AI market is actually headed. Both companies are building enterprise sales forces at a pace that would make a traditional software company dizzy, while simultaneously cutting private-equit...
image from Gemini Imagen 4
OpenAI and Anthropic are simultaneously scaling enterprise sales organizations at unprecedented speed, with OpenAI growing its GTM team from under 10 to 500 people in two years while planning to double total headcount to 8,000 by end of 2026. Both companies are raising billions from private equity firms at guaranteed 17.5% minimum returns through nearly identical dedicated enterprise JV structures, signaling a strategic bet that the enterprise AI adoption wave is still in early innings. The parallel moves suggest both labs are transitioning from research organizations to full-stack enterprise software companies with matching cost structure assumptions around continued high growth.
OpenAI and Anthropic are running the same play at the same time, and it tells you something about where the AI market is actually headed.
Both companies are building enterprise sales forces at a pace that would make a traditional software company dizzy, while simultaneously cutting private-equity checks that look less like desperation and more like a deliberate bet on distribution infrastructure. The question isn't whether the inbound pipeline will dry up. The question is what kind of company each lab is building toward — and whether the cost structure makes sense at 30 percent growth instead of 200 percent.
OpenAI's go-to-market build is the more documented of the two. According to Maggie Hott, OpenAI's head of enterprise sales, the entire go-to-market team was fewer than 10 people when she joined, focused entirely on API sales. No SDRs, no customer success managers, no working Salesforce instance. Two years later, that organization is 500 people strong. OpenAI is now planning to grow total headcount from roughly 4,500 to about 8,000 by the end of 2026, Bloomberg and the Financial Times reported. That is not a company preparing for a demand slowdown. That is a company betting everything on continued enterprise adoption.
Anthropic is running a parallel play, with different numbers but the same directional thesis. The company is targeting $20 billion to $26 billion in annualized revenue for 2026, Reuters reported, up from an annualized run rate approaching $7 billion in October 2025 and on track for $9 billion by year-end. Those are not the numbers of a company with an inbound problem. They are the numbers of a company that believes the enterprise AI buying wave is still in its early innings — and is trying to build the sales force before the window closes.
The most revealing part of both strategies is the private-equity maneuver. OpenAI is offering PE firms preferred equity stakes with a guaranteed minimum return of 17.5 percent, Reuters reported, and is in advanced talks with firms including TPG, Bain Capital, Advent International, and Brookfield Asset Management to raise about $4 billion at a pre-money valuation of roughly $10 billion for a dedicated enterprise JV. Anthropic is pursuing a similar structure with Blackstone, Hellman & Friedman, and Permira. Two labs, same deal architecture, same institutional counterparties. That is not coincidence. That is two companies looking at the same market and deciding that building enterprise distribution through a PE-partnered entity is more capital-efficient than scaling a direct sales force fast enough to capture the moment.
The "order-takers" framing that has been circulating around this story mistakes the symptom for the disease. Yes, both companies are hiring enterprise reps into a market where demand is currently strong enough that many accounts come inbound. But the PE strategy reveals the actual thesis: these labs are trying to build distribution infrastructure ahead of consolidation, not react to demand. They want channel partners, not just salespeople. They want PE capital to fund the build without loading the balance sheet, and they want relationships with the firms that advise large enterprises on technology strategy. The guaranteed return is the price of getting that distribution into the room.
The risk is not the inbound stopping. The risk is what happens to this cost structure when growth decelerates from 200 percent to something closer to the 30 percent that most enterprise software companies consider a strong year. OpenAI had roughly 4,500 employees as of early 2026 and is adding 3,500 more. Anthropic's headcount is not publicly disclosed but is widely estimated to be scaling at a comparable rate. Both companies are carrying sales organizations built for a hockey stick curve. The hockey stick may hold. But the PE partners are demanding 17.5 percent preferred returns regardless.
Enterprise penetration numbers give the build some cover. An a16z survey of 100 chief information officers, published in February 2026, found 78 percent already using OpenAI in production and 44 percent using Anthropic, with Anthropic's enterprise penetration rising 25 percent since May 2025. The same survey found 81 percent of companies now using three or more model families. (a16z discloses that it is an investor in OpenAI, which purchased a stake in Anthropic in March 2025.) That multi-model pattern is both opportunity and risk: it means the enterprise AI budget is real and expanding, but it also means neither lab has won a durable category. The customer is evaluating, not committed.
Claude Code, Anthropic's code generation tool, reached an annualized revenue run rate of nearly $1 billion in roughly six months, according to Reuters. That is the fastest product to $1 billion in ARR in the company's history, and it suggests the enterprise AI spending is not theoretical. But it also illustrates the tension: $1 billion in new ARR requires infrastructure to support, and Anthropic is simultaneously trying to hire enough enterprise sales capacity to go after the next $10 billion.
The framing that matters is infrastructure, not inbound. Both labs are building distribution the way a utility builds pipes — ahead of demand, at significant fixed cost, with the assumption that utilization will eventually justify the investment. That is a reasonable bet if enterprise AI adoption continues at current rates. It is a very different bet if the adoption curve flattens the way mobile did for some companies, or the way cloud did for others, before eventually accelerating again.
The PE firms taking the other side of this trade are not naive. A 17.5 percent guaranteed return on a $4 billion enterprise JV is not venture capital economics. It is private equity economics — predictable cash flows, portfolio company discipline, an exit strategy that probably involves the lab's IPO or a strategic acquisition. TPG, Bain, Advent, and Brookfield have collectively done thousands of enterprise software deals. They are not funding a moonshot. They are funding a sales and distribution infrastructure play that they believe will generate steady returns whether or not the AI market grows at 200 percent or 30 percent next year.
What to watch: whether the PE structures hold if growth decelerates materially, and whether the enterprise sales builds produce the renewal rates that justify the headcount. The labs have made their bet. The PE partners have made theirs. The enterprise customer, for now, is the one deciding whether both wagers were right.
Story entered the newsroom
Research completed — 0 sources registered. Two primary sources traced: (1) Reuters Oct 15 2025 by Krystal Hu for Anthropic 20-26B target - two unnamed sources, $7B run rate Oct 2025, targeting
Draft (1029 words)
Reporter revised draft (1030 words)
Reporter revised draft based on fact-check feedback
Approved for publication
Headline selected: OpenAI builds enterprise sales force as AI market shifts
Published (1048 words)
Get the best frontier systems analysis delivered weekly. No spam, no fluff.
Artificial Intelligence · 4h 1m ago · 3 min read
Artificial Intelligence · 4h 5m ago · 3 min read
