Palantir billionaire Alex Karp blasts AI industry as effing insane in fiery CNBC interview

Palantir Technologies chief executive Alex Karp called the artificial intelligence industry "effing insane" during a combative CNBC interview on Wednesday, accusing the biggest AI companies of overcharging their customers, harvesting valuable business data and putting U.S. national security at risk.

The billionaire cofounder, whose fortune Forbes pegs at $12.3 billion, appeared on the network to discuss a new partnership between Palantir and chipmaker Nvidia. The deal is meant to help U.S. government agencies deploy advanced AI more securely, and Karp cast it as an answer to mounting frustration among corporate leaders.

Karp said chief executives he speaks with privately are "livid" with the leading AI developers. He argued that companies such as OpenAI and Anthropic charge steep fees for their tools while collecting proprietary data that can be used to sharpen their own models, a practice he likened to a "wealth tax" on American business.

His sharpest words were aimed at the country's reliance on Silicon Valley firms for military and defense technology. "Are we really going to outsource the battlefield of this country to the consensus view in Silicon Valley?" he said. "That is effing insane."

When one host noted that he sounded angry, Karp did not back down. "This is the voice of American business that is being channeled through me," he said, adding that other executives would say the same in private.

The moment that drew the most attention came after the interview seemed to wrap up. Believing the segment had ended, Karp asked the hosts, "Are we still on?"

Investors appeared unbothered by the outburst. Palantir shares jumped more than 9% on Wednesday morning as the Nvidia tie-up landed with the market, extending a run that has made the company one of the most closely watched names in AI.

The partnership pairs Nvidia's open AI models with Palantir's software, which the two companies say will give government agencies and large enterprises more control over their data, more room to customize systems and lower costs than closed AI products allow. Karp has framed the pitch around ownership, telling customers they should control their own computing power, data and the models built on top of them rather than hand that value to outside providers.

The argument fits a case Karp has pressed for months. He has described model quality across the industry as converging into what he calls "commodity cognition," with the real advantage going to whoever owns the layer where AI is actually put to work. Palantir, he says, has more demand than it can supply.

He extended the same logic to the military. Karp said soldiers and U.S. allies deserve the best available technology but that many harbor trust issues with the frontier labs. Enterprises, he argued, feel they are paying for tokens that generate little value while surrendering the intellectual property and market edge that keep them competitive.

The company's numbers have backed the confidence. Palantir reported first-quarter revenue of $1.63 billion, up nearly 85% from a year earlier, and lifted its full-year guidance to about $7.65 billion. U.S. commercial revenue more than doubled. The stock, however, has had a rougher year, sliding roughly a third from where it started 2026 even after Wednesday's pop, as some investors balk at a valuation that trades far above peers.

Karp cofounded Palantir in the early 2000s with venture capitalist Peter Thiel, worth $27.4 billion, and Stephen Cohen, worth $4.6 billion. He met Thiel at Stanford Law School, where he earned a law degree before taking a doctorate in social theory in Germany and spending years managing money. The company took an unusual route to the public markets, listing directly on the New York Stock Exchange in 2020 rather than through a traditional offering, and has since become a central player in defense software and government AI.

His criticism lands at a tense moment between Washington and the frontier AI labs. The Pentagon in March designated Anthropic a "supply chain risk" after the company declined to lift restrictions that block its technology from being used for mass domestic surveillance or fully autonomous weapons. Around the same time, the Pentagon struck a deal with OpenAI that drew objections from AI policy and legal experts.

The federal government has moved to tighten its grip on the technology. President Donald Trump signed an executive order in June asking companies to allow federal review of new AI models before they reach the public. OpenAI said last week that wider access to its newest models would follow a limited preview restricted to a small group of trusted partners cleared by the government.

The pressure has not run in one direction. Anthropic said late Tuesday that the Commerce Department had lifted export controls on its newest models, Claude Fable 5 and Mythos 5, after the government had barred the company from letting foreign nationals use them over national security concerns. Commerce Secretary Howard Lutnick said officials had worked with Anthropic to review and improve the models and to strengthen U.S. leadership in AI.

Karp, known for shareholder letters that read more like political essays and for earnings calls delivered from rural New England, has rarely shied from provocation. His performance on Wednesday, national security warnings and profanity included, kept to that pattern while handing his company another jump in its share price.