Alphabet is guaranteeing lease payments on five US data centers that Anthropic will operate under a $35 billion financing arrangement, marking the first time a hyperscaler has backstopped real-estate obligations for a portfolio company rather than writing another equity round. The facilities have not been named, but the structure lets Anthropic add 400-500 megawatts of AI compute capacity without booking the liability or tying up cash that would otherwise fund model development.
Google already owns roughly 10 percent of Anthropic through previous equity injections totaling $2 billion and cloud-credit commitments that stretch into 2026. This lease guarantee is different. It converts Alphabet's strategic interest into a credit instrument, insulating Anthropic from the balance-sheet drag of long-term data-center leases while ensuring the facilities stay online even if revenue from Claude subscriptions or API calls falls short. The five sites are likely colocation builds with 18-24 month delivery windows, which means the first megawatt-hours should come online in mid-2026 and the last in early 2027.
The timing matters because OpenAI is closing a $40 billion credit facility with SoftBank and others to fund similar infrastructure, and Microsoft has already committed $80 billion in data-center capital expenditure for fiscal 2025. Anthropic was burning roughly $2.7 billion annually as of late 2024, with most of that spend going to compute rental from Google Cloud and Amazon Web Services. This financing lets the company lock in dedicated capacity without the front-loaded capital outlay that would force another dilutive equity raise or a down-round if the 2025 revenue targets—estimated at $850 million to $1 billion—slip.
For Alphabet, the guarantee solves two problems. First, it keeps Anthropic's training runs and inference workloads inside Google's ecosystem rather than migrating to AWS or Azure as the company scales. Second, it defers the need for another equity injection that would draw regulatory scrutiny in the US and UK, where antitrust authorities are already examining whether cloud credits and compute subsidies constitute control even when ownership stakes stay below 15 percent. The lease structure is harder to classify as a control investment, though the FTC will likely ask whether backstopping $35 billion in obligations creates the same lock-in.
The five data centers will likely house a mix of Nvidia H200 clusters for training and inference and Google's own TPU v5 pods, which Anthropic has used for earlier Claude iterations. The company is preparing to launch Claude 4 in the second half of 2025, a model that will require roughly 3-4 times the compute of Claude 3.5 Sonnet if parameter count and training data scale as expected. Without dedicated facilities, that training run would cost $600-800 million in rented compute, eating into the cash Anthropic raised in its $7.3 billion Series C last year.
Operators should watch whether Amazon, which also owns roughly 8 percent of Anthropic and supplies AWS Trainium chips under a separate agreement, matches this structure with its own lease guarantee. If it does, Anthropic will have effectively bifurcated its infrastructure between two hyperscalers, reducing single-vendor risk but complicating model deployment. The 2026 data-center delivery windows also overlap with the expected launch of Nvidia's Blackwell Ultra chips and potential export-control tightening on AI accelerators, either of which could delay or re-price the facilities.
The financing closes within 60 days, subject to due diligence on the underlying real estate and power-purchase agreements. If all five sites come online as planned, Anthropic will control more owned-equivalent compute than any AI lab outside OpenAI and xAI by the end of 2027.
The takeaway
Google converts strategic investment into infrastructure backstop, locking Anthropic into its cloud while avoiding equity dilution and regulatory flags.
Two hundred brands. Eight months on the desk. $0.003 an impression.
The branded-identity layer Chiefs of Staff and heritage CMOs route through — imprinting on real authorized stock for Nike, YETI, Patagonia, The North Face, Carhartt, Stanley, Peter Millar, TUMI, Montblanc, Moleskine, Waterford, and 190 more. Nine editorial desks publish the intelligence those operators read before they sign: The Stash Edge, Markets Edge, Sports Edge, Voyage Edge, Black's Edge, House Edge, the Article Engine, Ramen, and Fending.
$0.003per impression · vs ~$0.007 digital CPM
8 monthson the desk · vs 0.8s for a digital ad
200+authorized brands · Nike · YETI · Patagonia
9 deskspublishing daily · since 1997
70,000 SKUs · virtual proof in 60 seconds · no platform fee · blind-shipped · ASI #217876
Your next customer won't visit your website. Their AI will.
AI assistants have quietly taken over the first step of buying — they answer from catalogs they can read and shortlist whoever can actually ship. Two questions now decide whether you exist to that buyer: can a machine read your catalog, and can you fulfill the order. Most brands fail one or both and never find out why the orders went elsewhere. The winners of this shift aren't the loudest. They're the most readable. Build for the machine that's about to do the shopping.
Built by the craft floor — apparel, media, packaging, and secure print.
This trade runs on hands, not desks. Imprint manufacturing & Komori Press · Canon high-speed secure-media operations is a craft floor — genuine Six Sigma discipline applied to ink, thread, foil, and registration, where a hundredth of an inch is the difference between a brand that reads serious and one that reads cheap. POPS4 is built by exactly those operators: independent, boots-on-the-ground engineers who carry their own book, read a client in microseconds, and put their name on every run. Beyond our own Virginia Beach floor, we work with a vetted network of craft manufacturers across the US — each meeting the highest excellence in QC standards in the industry, each a specialist in its own discipline — so apparel, hard-goods imprinting, media manufacturing, packaging, and secure printing all go to the bench built for them, coordinated from one accountable hub. Short-run from twenty-five units, volume to five hundred thousand. Two hundred authorized national brands, seventy thousand SKUs with virtual proofing on every one. Art archived for instant reorders. Net-thirty corporate terms, NDA-standard white-label — your name on the work, or none at all.
Strategy, positioning, identity, creative, and messaging — wired into an AI system that publishes and distributes on its own. Nine editorial desks generate the authority, the production house ships the physical proof, and the attribution layer tells you which post sold which SKU. What you get is an operating layer — content, catalog, and order path under one roof — that keeps working whether or not you are in the room. Built for principals who would rather own the machine than rent the agency.
Named-account programs — one desk, quiet delivery, NDA-standard.
One point of contact who already knows the file, so nothing restarts from zero between engagements. The work ships blind, under NDA, with your name on it or none at all. Built for single-family offices, heritage-house CMOs, sports-ownership groups, and the agencies that white-label our production. The relationship is the product; the merch is the proof of it.
SFO · Chief of Staff desk. Principal household, properties, aircraft, yacht, calendar, philanthropy — one file.
Shop seventy thousand products. Virtual proof on every one. 24/7.
Drop your logo on any product and see the virtual proof before asking. Quote routes direct to the desk. MCP catalog for AI agents. Celeste for the fast conversation. Full self-service checkout in development.