Palantir’s Alex Karp warns AI hype is outpacing returns, targets OpenAI and Anthropic

Palantir CEO Alex Karp used a CNBC appearance this week to argue that the industry has been too quick to market AI as an immediate, universal solution, saying it has been irresponsibly oversold by prominent players.

In the segment, as reported by Yahoo Finance via 247WallSt, Karp singled out OpenAI and Anthropic for what he described as overselling the technology, while pointing to a broader pattern of companies spending heavily to use AI systems without sufficiently defined outcomes.

A central frustration in Karp’s critique is the economics of deploying AI at scale. The report characterizes his view as driven by concern that businesses are paying for access to AI capabilities, including costs associated with processing and tokens, while struggling to show durable returns that match the spend.

The exchange also reflects Palantir’s long-standing emphasis on enterprise deployments rather than purely model-centric hype. While the cited reporting does not lay out any new Palantir product announcements in connection with the remarks, Karp’s skepticism implicitly challenges a market pitch that treats AI usage itself as the value proposition.

Karp’s comments come at a time when enterprise AI budgets are being tested by operating realities. Even when models perform well in demos, companies still must integrate them into workflows, validate results, and manage ongoing costs. The report’s framing suggests Karp believes that some buyers are moving faster than the value case.

What “fix” Karp advocates is not fully detailed in the available reporting. The Yahoo Finance write-up, as republished by 247WallSt, indicates he believes there is a better way to approach AI adoption, but the specific mechanics, targets, or timelines of any proposed shift were not included in the material available for this review.

For Palantir, the risk in this debate is reputational as well as commercial. If enterprise customers conclude that the broader AI market is simply too expensive or too uncertain, that could compress budgets not only for competitors, but also for vendors selling AI-enabled platforms. The CEO’s critique of major model makers may also influence how buyers weigh competing approaches to enterprise AI.

Still, the remarks can also be read as a demand announcement: enterprise buyers want business-linked outcomes, not just access to powerful models. Investors will likely watch whether Palantir’s own messaging and customer traction continues to align with that outcome-focused framing, particularly as public companies report AI spending and ROI in upcoming earnings cycles.