The land grab has gone financial
OpenAI's 17.5% guaranteed-return PE pitch, its 450,000 sq ft campus lease, and the Helion fusion deal all point to the same shift: the AI race is no longer about who has the best model — it's about who can lock in distribution, real estate, and energy before the IPO window opens. Anthropic is running the same playbook at smaller scale, and Musk is vertically integrating chips. The competition has moved from the lab to the balance sheet, and the winners will be decided by deal structure as much as by benchmark scores.
The Decoder
OpenAI guarantees private equity firms 17.5% returns to outflank Anthropic in enterprise AI race
OpenAI is offering TPG, Bain Capital, Advent International and Brookfield a guaranteed 17.5% minimum return plus early model access in a $10 billion joint venture designed to roll out AI across hundreds of portfolio companies — a direct counter to Anthropic's own, smaller PE pitch.
the-decoder.com

The AI race stopped being about models weeks ago. It might have stopped months ago, but this week made it obvious. The competition between OpenAI, Anthropic, and everyone else has become a financial engineering contest, and the scoreboard is denominated in square footage, gigawatts, and guaranteed returns.
Consider what OpenAI announced in a single week. The Decoder reported that OpenAI is pitching TPG, Bain Capital, Advent International, and Brookfield on a $10 billion joint venture that would guarantee private equity firms a 17.5% minimum return plus early access to frontier models. The structure is remarkable: $4 billion in preferred equity at a ~$10 billion pre-money valuation, with a floor return that shifts risk back onto OpenAI. Anthropic is running a parallel play with Blackstone, Hellman & Friedman, and Permira, but at roughly $1 billion and without the guaranteed return. The gap in aggression tells you everything about who feels the IPO clock ticking loudest.
That PE deal isn't happening in isolation. BusinessWire reported that OpenAI signed a 10-year lease for 450,000 square feet of Class A office space in Mountain View, owned by KKR Real Estate Finance Trust. Five buildings. The entire campus. This is a company preparing to nearly double its headcount to 8,000 by year's end, and it's doing so by committing to a decade of physical presence in a market where most tech firms are still negotiating hybrid work policies.
Then there's the energy play. TechCrunch reported that OpenAI is negotiating to secure 5 GW of fusion power from Helion by 2030, scaling to 50 GW by 2035. Sam Altman stepped down as Helion's board chair to manage the conflict of interest, which tells you the deal is serious enough to require it. Helion's Polaris prototype recently hit 150 million degrees Celsius, and Microsoft signed a similar power purchase agreement back in 2023. The pattern is clear: frontier AI companies are locking in energy supply the same way airlines lock in fuel hedges, years before they need it, because being right and being late produce the same outcome.
The IPO shapes everything
All of these moves make more sense when you read them through the lens of Yahoo Finance's reporting that OpenAI is targeting a Q4 2026 IPO and has told staff that ChatGPT must be repositioned as a "productivity tool." The PE joint venture absorbs the messy, high-cost enterprise deployment work that would otherwise drag down segment reporting. The campus lease signals permanence to public market investors. The energy deal signals long-term thinking. Each piece is an IPO narrative in physical form.
The way I see it, the interesting question isn't whether these bets pay off for OpenAI. With annualised revenue past $25 billion, they probably do. The question is what happens to everyone else. When one player guarantees PE firms 17.5% returns to win distribution, the cost of competing goes up for every other AI company. Anthropic's smaller, no-guarantee PE pitch looks measured by comparison, but "measured" is another word for "outgunned" when the other side is writing cheques this size.
For builders, the implication is practical. The companies providing your inference aren't competing on benchmarks anymore. They're competing on who can secure the cheapest energy, the most distribution partners, and the most favourable capital structure before the public markets open. That financial layer will determine pricing, availability, and vendor lock-in for the next five years. The model is table stakes. The deal sheet is the moat.
Read the original on The Decoder
the-decoder.com