The British fintech founder Tom Blomfield is moving from Y Combinator to Anthropic and will work on the compute team on the computing capacity for Claude. He announced this himself on July 13, 2026, on the platform X. A few weeks earlier, Anthropic had completed a funding round of 65 billion dollars at a valuation of 965 billion dollars.
Blomfield moves to the compute team for computing capacity
According to his own statements, Blomfield will be a member of the technical staff and will work directly with Tom Brown, co-founder of Anthropic and the company’s Chief Compute Officer. He is taking a leave of absence from Y Combinator, where he had been a full partner since May 2023 and had previously advised founding teams in the fintech sector as a guest partner from November 2021. On X, he explained the move by saying that powerful AI has the potential to improve the life of every person on earth, and that the availability of computing capacity in the early stages of recursive self-improvement is becoming one of the most important open questions. The move follows a series of prominent external hires at Anthropic: in May 2026, former OpenAI co-founder Andrej Karpathy joined the pretraining team, and the company also brought in former Microsoft manager Eric Boyd for a leadership role. Anthropic is thus clearly favoring people with entrepreneurial and commercial experience over purely research-focused profiles. Brown himself co-founded Anthropic in 2021 after previously being involved in training the GPT-3 language model at OpenAI.
Anthropic secures capacity through several billion-dollar deals
Anthropic’s build-out of its computing infrastructure rests, as The Next Web reports, on several parallel agreements. The plan calls for deploying around one million Google TPUs, with more than one gigawatt of capacity expected online as early as 2026 – the underlying Google deal is said to run into the tens of billions of dollars, according to Tech Funding News. A separate agreement with Google and chipmaker Broadcom is set to add 3.5 gigawatts of next-generation TPU capacity starting in 2027. Through a partnership struck with xAI in May 2026, Anthropic also gains access to more than 220,000 Nvidia GPUs in xAI’s Colossus data centers, which were originally built as part of a partnership with SpaceX. British AI consultant Gail Weiner, speaking to PYMNTS, frames the hire as a sign that computing capacity has long since stopped being a purely technical problem and has become a commercial and operational one as well. For negotiations with chip and energy suppliers, dealmakers like Blomfield appear better suited than traditional infrastructure engineers, since it comes down to prepayments, delivery commitments, and capacity reservations in a race against Google, Microsoft, and OpenAI, in which data center space and power hookups are just as scarce as the chips themselves.
Blomfield brings experience from two billion-dollar exits
Blomfield founded the British mobile bank Monzo in 2015 and served as its CEO until 2020; back in 2011, he had already co-founded the payments provider GoCardless. GoCardless has since been sold to payments company Mollie for 1.05 billion euros, and the combined peak valuation of both companies reportedly topped nine billion dollars. Anthropic itself announced in its own statement in May 2026 a Series H round of 65 billion dollars at a valuation of 965 billion dollars, alongside a run-rate revenue of more than 47 billion dollars and an investor group including Altimeter Capital, Dragoneer, Greenoaks, and Sequoia. With that valuation, Anthropic overtook OpenAI as the most valuable pure-play AI company, according to CNBC – how solid that direct comparison between two privately held companies really is remains independently unverified. The company is targeting an IPO for October 2026, though no official details on the terms have been released; Monzo, meanwhile, is separately preparing its own London listing, with an expected valuation of six to seven billion pounds.
What will matter is whether Anthropic’s bet on commercial dealmaking over pure engineering capacity pays off: the announced gigawatt-scale capacity still has to actually come online, while competition for chips, power, and data center space among the major AI providers keeps intensifying.
