Chamath Palihapitiya says consulting giants are serving to OpenAI and Anthropic construct the companies which will substitute them.
Abstract
- Chamath Palihapitiya says consulting giants like PwC and Accenture are endangering their very own companies by instantly embedding OpenAI or Anthropic instruments with out sustaining management over token era.
- OpenAI launched the OpenAI Deployment Firm with over $4 billion in funding at a $10 billion valuation, whereas Anthropic raised $1.5 billion for a rival enterprise providers enterprise — each instantly competing with conventional consulting companies.
- Palihapitiya’s personal agency, 8090, has partnered with EY on an AI-native software program supply platform known as EY.ai PDLC, which he says offers consultants the “management aircraft” wanted to route AI tokens to any mannequin supplier.
Billionaire enterprise capitalist Chamath Palihapitiya issued a stark warning to the consulting business on Could 17, calling out PwC and Accenture by title for what he describes as a self-defeating AI technique — one which he says arms OpenAI and Anthropic the keys to their very own substitute.
“In case you are working a consulting enterprise and you might be deploying Anthropic or OpenAI instantly into your group (I’m taking a look at you PwC and Accenture) you might be letting the fox into the hen home,” Palihapitiya wrote on X. “OpenAI and Anthropic are brazenly funding and beginning opponents to you whereas additionally utilizing your utilization to drive extra success for them.”
The put up got here days after OpenAI formally launched the OpenAI Deployment Firm, a subsidiary backed by over $4 billion from 19 institutional buyers — together with Goldman Sachs, TPG, Brookfield, Bain Capital, McKinsey & Firm, SoftBank, and Capgemini — at a $10 billion pre-money valuation. Concurrently, as crypto.information reported, Anthropic launched a competing enterprise providers agency backed by $1.5 billion from Blackstone, Hellman & Friedman, and Goldman Sachs. Collectively, the 2 labs dedicated $5.5 billion to construct out consulting and implementation capabilities that place them in direct competitors with the companies presently deploying their fashions.
Token management by way of OpenAI and Anthropic as the brand new digital moats
Palihapitiya’s argument facilities on what he frames as a structural vulnerability: consulting companies that pipe shopper workflows instantly by means of OpenAI or Anthropic fashions are permitting those self same labs to be taught from utilization patterns, construct out competing service arms, and finally displace the companies appearing as middlemen. “This isn’t a failure on their half however a failure in your half,” he wrote.
His proposed various is a “management aircraft” that lets consulting companies arbitrate which tokens go the place and which mannequin supplier generates them. “Controlling the tokens is controlling the spice,” Palihapitiya wrote, borrowing a reference from Dune. He cited his personal software program enterprise, 8090, for instance of this method in follow, pointing to the corporate’s international partnership with EY, introduced in March 2026. Beneath that deal, EY adopted 8090’s Software program Manufacturing facility platform to energy EY.ai PDLC, an AI-native product growth lifecycle software projected to extend software program growth productiveness by 70% and speed up supply by 80x, in response to EY’s personal figures. The platform is being rolled out to tens of hundreds of EY US consultants. “We management token era and may direct them to any mannequin supplier,” Palihapitiya wrote, including that 8090 is “shut to a different international partnership” to be introduced quickly.
The stakes for legacy consulting
The aggressive risk Palihapitiya describes is already materializing in income information. As crypto.information reported, Anthropic’s run-rate income hit $30 billion in early April 2026, tripling from $9 billion at year-end 2025, with over 1,000 enterprise prospects every spending greater than $1 million yearly. OpenAI, in the meantime, crossed $25 billion in annualized income as of February 2026, as crypto.information coated, with enterprise income already accounting for over 40% of whole earnings. Each labs are actually buying engineering expertise and constructing forward-deployed engineer groups modeled on Palantir’s playbook — embedding specialists instantly inside shopper organizations, the very function that companies like Accenture and PwC have traditionally owned.
Palihapitiya closed with a pointed characterization of companies sluggish to adapt: “These organizations refuse to simply accept the disruption standing nonetheless or, even worse, by adopting and accelerating the businesses who wish to disrupt them.” For conventional consultancies navigating a quickly shifting enterprise AI panorama, the warning alerts that the window to construct model-agnostic infrastructure — moderately than merely reselling AI entry — could also be closing sooner than many had anticipated.