Palantir shares jump 9% as Karp blasts AI labs' token pricing model

Palantir Chief Executive Officer Alex Karp launched a blistering attack on OpenAI and Anthropic's token-based pricing model, calling it a "wealth tax" that forces enterprises to pay rising costs while handing over their most valuable data and intellectual property.

"Something has gone completely wrong," Karp said in a CNBC interview Wednesday. "The basic view among enterprises in this country is I'm going to chillax and waste my time with tokens, I'm going to get no value, and they're going to get my IP."

Karp's criticism comes as enterprise customers grow increasingly frustrated with soaring AI costs. The so-called "tokenmaxxing" era — where companies spent aggressively on AI tokens without clear returns — is giving way to a focus on efficiency and measurable return on investment. Palantir shares rose 9 percent on Wednesday as investors embraced the company's alternative vision.

The debate over AI pricing models carries significant implications for the $200 billion-plus enterprise AI market. If enterprise customers shift toward open-weight models and on-premise deployments — where Palantir and Nvidia are positioning themselves — it could pressure the valuation of private AI labs like OpenAI, reportedly valued at $300 billion, and Anthropic, valued at $60 billion.

The 'AI Sovereignty' Manifesto

A day before the interview, Palantir published a nine-point "AI Sovereignty" manifesto on X, laying out its critique of the current AI market. The document argues that institutions ceding control of their data, model weights, and competitive advantage to third-party AI labs puts their future at risk. "Data retention is your treasure. Transfer it at your own peril," the manifesto states, while decrying the "addictive sense of false progress" that token-based pricing creates.

Karp framed the issue in stark national security terms. "Are we really going to outsource the battlefield of this country to the consensus view in Silicon Valley? That is effing insane," he said, appearing to reference Anthropic's restrictions on military use of its models.

Open Models Gain Ground

Karp's critique aligns with a broader industry shift. Earlier this week, Palantir announced an expanded partnership with Nvidia to build custom AI models for U.S. government agencies using the chipmaker's AI tools. Karp emphasized that enterprises want "control over their compute, their models, their data stack and their alpha."

"What aligns me with Nvidia, and I think is what the technical customers want, which is control over their compute, their models, their data stack and their alpha," Karp said. "They want to know they own the means of production. It's not being transferred to someone else."

The push toward open-weight models is also being fueled by Chinese AI labs, which are rapidly closing the capability gap with U.S. frontier models at a fraction of the cost. Karp warned the industry not to underestimate the speed of China's progress.

For investors, the growing enterprise backlash against token pricing creates a clear competitive dynamic. Palantir, trading at about 60 times forward earnings, is positioning itself as the enterprise-friendly alternative to the cloud-based AI labs. Nvidia stands to benefit as more enterprises seek on-premise AI infrastructure. Meanwhile, OpenAI and Anthropic face mounting pressure to justify their pricing models as customers demand tangible returns on their AI spending.

This article is for informational purposes only and does not constitute investment advice.