Capital & Industrial Strategy

Anthropic has formally unveiled a $1.5 billion joint venture involving Blackstone, Goldman Sachs, and Hellman & Friedman, with each firm contributing approximately $300 million. The new entity — structured as a consulting company — will advise financial services firms on deploying Anthropic's AI across investment portfolios. This is a confirmed deal, not a proposal. Financial Times, Wall Street Journal, and CNBC all confirm the structure. The strategic logic is vertical integration into the client relationship: rather than selling API access to financial firms and competing on price, Anthropic is co-owning the deployment infrastructure and capturing a share of the value created. Blackstone and Goldman are also natural distribution channels into the PE-owned portfolio companies that represent the JV's primary target market.

Simultaneously, OpenAI is executing a parallel enterprise push — also structured around partnerships with asset managers — though terms and capital commitments for OpenAI's arrangement are less clearly confirmed in current reporting. TechCrunch and Semafor report both companies are moving in this direction. The competitive dynamic is instructive: both frontier model labs are independently concluding that raw API distribution is insufficient to capture enterprise value at scale, and are instead racing to build proprietary go-to-market infrastructure with financial sector gatekeepers.

Meta is assembling a roughly $13 billion financing package — led by Morgan Stanley and JPMorgan — to fund a data center in El Paso, Texas. This is a confirmed arrangement, per Bloomberg. The use of structured debt for AI infrastructure marks a material shift in how hyperscalers are funding capex: rather than deploying equity off the balance sheet, Meta is treating data center assets as financeable infrastructure with predictable cash flows — a move that mirrors project finance logic from energy or real estate. Morgan Stanley Co-President Dan Simkowitz, speaking at Milken, confirmed AI financing as a core theme driving investment banking activity. Bloomberg

At the more speculative frontier, Peter Thiel has led a $140 million investment into Panthalassa, an ocean-based data center startup powered by wave energy, as part of a $1 billion raise. Financial Times reports this as a confirmed close. The strategic driver is power availability: as grid-connected land sites face permitting and energy constraints, capital is flowing toward unconventional infrastructure solutions. SK Hynix shares rallied 13% on continued AI spending signals from US tech firms, per Reuters, reflecting the supply chain read-through from hyperscaler commitments.

Cerebras Systems is targeting a $3.5 billion raise in its IPO, with a valuation of up to $26.6 billion, per TechCrunch and CNBC. Its deep commercial relationship with OpenAI — providing inference compute — is central to its growth narrative. The IPO will be a significant market signal: if Cerebras prices at the top of its range, it validates investor appetite for Nvidia alternatives in the inference stack. Note that valuation figures differ slightly across sources ($24.5 billion per CNBC vs. $26.6 billion per TechCrunch), likely reflecting different dilution assumptions; the range rather than a single figure should be the reference point.

At the application layer, Sierra raised $950 million to pursue what it describes as the global standard for AI-powered customer experience, per TechCrunch. The scale of the raise — giving Sierra over $1 billion in deployable capital — reflects investor willingness to fund horizontal enterprise AI platforms at significant pre-revenue-scale valuations. Separately, Nvidia backed DeepInfra's $107 million Series B, a cloud inference platform targeting compute bottlenecks, with Samsung also participating per Bloomberg. Nvidia's strategic rationale is clear: backing inference platforms that run on its hardware locks in demand and provides competitive intelligence on where inference bottlenecks are emerging.

Davidson Kempner CIO Tony Yoseloff has publicly stated that AI disruption is threatening recovery rates for private credit firms with exposure to the software sector, per Bloomberg. This is a significant development: Davidson Kempner is a major distressed-debt and special-situations investor, and its CIO's public commentary signals internal reassessment of enterprise software collateral valuations in existing portfolios. The mechanism is straightforward — if AI commoditizes software functions previously monetized by SaaS incumbents, the revenue bases underpinning leveraged loans deteriorate, reducing lenders' recovery in default scenarios.

This risk is not theoretical. The private credit market has extended substantial leverage to software companies — many PE-owned — at valuations premised on stable recurring revenue growth. If AI-native alternatives erode those revenue streams faster than amortization schedules, lenders holding junior debt face structural impairment. The Anthropic-Blackstone-Goldman JV targeting PE-owned portfolio companies is partly a defensive play by those same financial institutions: by helping portfolio companies adopt AI, they protect the revenue bases that backstop their credit positions.

The Wall Street Journal reports that OpenAI has internally discussed spinning out its robotics and hardware divisions into separate entities, potentially adopting an Alphabet-like holding structure — though the WSJ notes no discussions are currently active. WSJ This is speculative intent, not a confirmed plan. The strategic logic is IPO readiness: a cleaner corporate structure with a pure-play AI model business at the core would be more legible to public market investors than a conglomerate spanning models, chips, robotics, and consumer products.

In the Musk v. Altman federal trial, OpenAI president Greg Brockman disclosed a stake valued at approximately $30 billion in the context of the company's for-profit restructuring, per Wired and FT. Brockman's testimony framing his equity as a reward for founding contribution is legally defensive — the Musk lawsuit alleges executives are enriching themselves by abandoning the nonprofit mission. The disclosure matters to capital market observers because it crystallizes the magnitude of personal financial stakes among OpenAI's inner circle, which helps explain executive incentive alignment around the restructuring and IPO timeline.