In a direct interview, Karp argued that many large companies are paying huge AI bills without realizing significant business value. He claimed that companies are charged for tokens rather than results, raising a simple question: If AI really can generate billions in value, why aren't vendors charging a portion of that value instead of billing by computation?

Karp also warned that companies risk losing their competitive edge by feeding sensitive data, workflows, and insider knowledge into frontier AI systems. In his view, companies could end up helping train the same models that power products for their competitors.

He described the current model as a "wealth tax" on companies, noting that executives privately worry about high costs, weak return on investment and the protection of their intellectual property, even though few are willing to publicly criticize large AI labs.

Whether or not his claims are true, Karp's comments highlight a growing debate around AI: Are companies buying transformative business value, or simply paying ever-larger compute bills while delivering valuable data?