Fiercely criticizing big AI model companies! Palantir CEO: Their business model has huge problems, imposing a "tax" on enterprises without regard for corporate risks.

```

Palantir CEO Alex Karp launched a fierce attack against leading AI labs such as OpenAI and Anthropic, pointing out fundamental flaws in their business models and warning that enterprise clients are paying rising token costs for limited value, while putting their core data and intellectual property at risk.

On Wednesday, in a CNBC interview, Karp used strong language, calling the business models of top AI labs "utterly crazy," and accused Anthropic and OpenAI of misleading enterprise partners—exaggerating AI risks while selling their most powerful models to global corporations and governments.

He characterized this phenomenon as enterprises being forced to pay a "wealth tax," essentially surrendering their competitive edge to third parties. When CNBC host Becky Quick said "you sound very angry," Karp replied: "This is the voice of American business, conveyed through me."

These remarks reflect growing dissatisfaction among enterprise clients with US AI labs. High costs, questionable return on investment, and increasingly tight regulatory pressures have driven some large American corporations to turn to more affordable Chinese alternatives, putting institutions like OpenAI and Anthropic under multiple pressures.

"Something is fundamentally wrong"

Karp directly targeted the token-based billing models used by Anthropic and OpenAI. "I'm not targeting them, but something is fundamentally wrong," he said. "Enterprises in this country generally believe I'm just wasting time and resources with tokens."

He believes that while AI labs exaggerate risks, they aggressively promote powerful models, and enterprise clients are in fact paying a "wealth tax" to these platforms, transferring their "alpha"—their core competitive advantage—to external parties. On national security, his words were even stronger: "Are we really going to outsource this country's battlefield to Silicon Valley's mainstream consensus? That's just crazy."

Nine-Point "AI Sovereignty" Manifesto

Karp's public statements are not isolated. Just the day before the interview, Palantir released a nine-point "AI Sovereignty" manifesto on X, systematically outlining its critical stance toward the current AI ecosystem.

The manifesto centers on the logic of "sovereignty": If institutions relinquish control over their data, model weights, and competitive advantages, they are surrendering their future choices to others, who will likely profit from them. The manifesto explicitly states that "tokenmaxxing" (token maximization) incentives lead to one-off scripts instead of robust software systems and create an "addiction to false progress"—"Those who sell tokens refuse to charge by value, and there are reasons for that."

The manifesto also emphasizes that controlling model weights means controlling the fate of the institution. "Weights are the essence of the knowledge painstakingly accumulated by the institution. If you let others control your weights, you allow them to migrate your business alpha into their hands."

A Microcosm of the Corporate 'Rebellion' Wave

Karp's criticism is not an isolated incident but a concentrated reflection of broader dissatisfaction among enterprise clients toward US AI labs. High usage costs, hard-to-quantify real returns, and increasingly complex regulatory environments are jointly eroding enterprise trust in mainstream AI platforms.

Some large American companies have begun turning to more cost-effective Chinese AI alternatives, directly pressuring institutions such as OpenAI and Anthropic. Palantir’s stance is clear: companies should treat data as a core asset and see AI sovereignty as a precondition for institutional survival, rather than entrusting their fate to external platforms' "consensus judgments."

Risk Warning and DisclaimerThe market has risks, and investment needs caution. This article does not constitute individual investment advice, nor does it take into account the unique investment goals, financial situation, or needs of any particular user. Users should consider whether any opinions, viewpoints, or conclusions in this article suit their specific circumstances. Investment is at your own risk. ```