‘Something has gone completely wrong’: Palantir CEO Alex Karp targets OpenAI, Anthropic over AI pricing

Palantir CEO Alex Karp has criticised the token-based pricing model used by OpenAI and Anthropic, arguing that many businesses are spending heavily on artificial intelligence without seeing meaningful returns.Speaking to CNBC, Karp said the way frontier AI companies sell their products has left enterprise customers increasingly dissatisfied. While stressing that his criticism was not directed personally at OpenAI CEO Sam Altman or Anthropic co-founder Dario Amodei, he questioned the industry's current business model."So the general way these things were sold... Sam and Dario, there's nothing more fun than debating Dario in private. So I'm not throwing shade at them, but something has gone completely wrong," Karp said.Most leading AI companies charge customers based on token usage, with tokens serving as the units of text processed by AI models. Karp argued that the pricing model is failing to generate enough business value for enterprise customers."And 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,'" he said, referring to concerns that businesses are paying for AI services while exposing valuable intellectual property.Karp said frustration among corporate customers has continued to grow."Every single enterprise I deal with... these people are livid. They're like, 'I am paying for tokens that create no value. These people are stealing the weights and alpha of my business,'" he said.Palantir, which develops AI software for governments, defence agencies and large enterprises, has positioned itself as an alternative that allows organisations to retain greater control over their data and AI systems instead of depending entirely on model developers.Karp pushes for greater AI controlBeyond pricing, Karp also questioned whether private AI companies should retain control over models used in sensitive government and military applications."In the classified context, when the Department of War goes to you and says, 'I need this application,' do they get to control the weights to do it? Or do you get to control the weights? Are we really going to outsource the battlefield of this country to the consensus view in Silicon Valley? That is effing insane," he said.Karp argued that governments and organisations deploying AI should have authority over model weights, the internal parameters that determine how AI systems behave and can be adapted for specialised tasks.He also said businesses increasingly want greater ownership of the technology powering their AI deployments instead of relying completely on commercial AI providers."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."Karp also cautioned against underestimating China's progress in artificial intelligence and said more enterprises are building proprietary AI systems tailored to specific business needs rather than relying solely on large general-purpose models.Palantir also recently released a nine-point "AI sovereignty" manifesto, criticising what it described as "tokenmaxxing" and encouraging organisations to retain ownership of their data.Karp said his comments reflected the concerns he hears from businesses across the United States, adding that many enterprises are tired of what they believe has been an overselling of AI capabilities.Read More: OpenAI explores giving US government a 5% stake: Report