Palantir renews push for enterprise AI control as Karp attacks frontier labs
As companies weigh how to deploy artificial intelligence across core operations, Palantir is sharpening its case for keeping data and model control inside the enterprise. CEO Alex Karp says businesses should not have to surrender intellectual property to work with model makers, tying that argument to Palantir's latest earnings momentum.
Highlights
- Palantir CEO Alex Karp warns on Monday that frontier AI developers are pressuring enterprises to cede control of their data and intellectual property.
- Karp advocates that businesses should control their AI models and keep data in-house, as echoed by concerns from major government clients.
- Palantir reports overall revenue up 93% and U.S. commercial revenue up nearly 150% year-over-year in Q2, emphasizing growth with enterprise-controlled AI solutions.
AI sovereignty strategy and earnings backdrop
As reported by CNBC, Karp intensifies his criticism of frontier AI developers on Monday, saying enterprises are being pushed toward a model in which they lose control over their own assets and future competitiveness. In an interview, he argues that companies should not be compelled to hand over intellectual property in exchange for access to advanced models.Karp says businesses instead should control their own models and use companies such as Palantir for an application layer that sits on top of an existing corporate technology stack. That approach, he says, allows organizations to keep their data in-house while still advancing artificial intelligence use across operations.
He also says Palantir hears similar concerns from clients, including major government enterprises, which question paying for tools they do not control and that may not preserve the value created inside their businesses.
Implications for U.S. enterprise AI competition
Karp links that message to what he describes as a broader global movement around AI sovereignty, a theme he repeats in his second-quarter letter to investors released Monday alongside Palantir's earnings. In that letter, he writes that organizations are increasingly recognizing the risks of giving language model creators deep access to their institutions.The comments come as Palantir reports strong financial results. The company says overall revenue grows 93% from a year earlier, while its U.S. commercial business jumps nearly 150% in the second quarter, reinforcing its effort to position enterprise-controlled AI as both a strategic and commercial alternative to frontier model providers.
In our earlier article on Palantir’s upgraded full-year revenue outlook after its second-quarter results, we noted that demand for its U.S.-built AI software helped the company beat expectations and raise guidance. We also highlighted how the AI sovereignty narrative is boosting domestic commercial momentum while international revenues soften as some foreign governments look to reduce reliance on U.S. tech providers.