Palantir CEO: U.S. AI Labs Losing Ground to Chinese Models
Alex Karp says enterprise customers are abandoning expensive American AI tools for cheaper alternatives, as regulatory pressure mounts.
Palantir CEO Alex Karp delivered a sharp rebuke of leading U.S. AI companies during a CNBC appearance Wednesday, arguing that OpenAI and Anthropic have prioritized building powerful models over serving enterprise customers — and are now paying the price.
According to Karp, CEOs are privately expressing frustration that they're getting "no value" from enterprise AI tools despite steep costs. That dissatisfaction is driving a significant shift: American companies are increasingly adopting Chinese AI models to reduce expenses, even as the Trump administration restricts access to some domestic AI capabilities.
Why it matters
The combination of high costs and regulatory uncertainty is creating an opening for foreign competitors at a critical moment in the AI race. If U.S. enterprises can't justify the expense of American AI tools or face access restrictions, the domestic AI industry risks losing commercial momentum to international rivals — particularly China.
The migration to Chinese models
The cost pressure is real and measurable. Microsoft is reportedly considering DeepSeek, a Chinese AI model. Coinbase has kept AI expenses flat by incorporating Chinese open-weight models. U.S. startup Cursor built its latest offering on Kimi 2.5, developed by Moonshot AI with backing from Alibaba.
Data from OpenRouter confirms the trend: usage of Chinese AI models is surging as enterprises hunt for ways to lower their AI bills, according to details first reported by Axios.
Regulatory friction compounds the problem
Karp also criticized how AI risks have been communicated, saying models were "irresponsibly" labeled as "dangerous for everyone" without clear guidance on protecting enterprise intellectual property. This messaging has contributed to regulatory headwinds.
The White House has asked both Anthropic and OpenAI to limit or delay releases of their most advanced models. Some developers are now blaming Anthropic CEO Dario Amodei for triggering Washington's response through vocal warnings about AI risks. One developer told Axios they're reducing their use of Anthropic's tools partly because of Amodei's cautionary statements.
The competitive landscape shifts
Karp's critique comes as Palantir itself competes with the companies he's criticizing. While Palantir doesn't train frontier models, it increasingly vies with OpenAI and Anthropic to become the platform through which enterprises and governments deploy AI.
Despite his criticism, Karp acknowledged Amodei as a "historic figure" and praised Anthropic as one of the fastest-growing companies in American history. He also noted that the AI race has been characterized by constant leapfrogging, suggesting current leaders shouldn't be counted out.
Still, the message is clear: after years of hype and apparent invincibility, U.S. AI labs are facing real pushback from major customers over both economics and regulation. Whether this represents a temporary setback or a more fundamental competitive challenge remains to be seen.
These details were first reported by Axios.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call
