Karp Slams AI Token Model as Palantir Bets on Sovereign Compute

Palantir CEO Alex Karp criticizes OpenAI and Anthropic’s token-based AI model as unsustainable for enterprises, advocating ROI-driven, open-weight approaches and data ownership. Palantir posted strong AI-driven results: Q1 FY2026 revenue of $1.63B (+84.7% YoY), US commercial revenue up 133% to $595M, and adjusted operating margin of 60%, with annual growth guidance raised to 71%. The company also expanded its Nvidia partnership to build sovereign AI models for government and enterprises, while investors like Michael Burry added Palantir puts. Despite the growth, Palantir trades around a forward 91x earnings, and its RDV sits at $4.92B as it pushes the “own the means of production” thesis amid cost pressures in AI tooling (e.g., Uber’s AI-spend cap).
- Palantir CEO: "Something Has Gone Completely Wrong" With OpenAI and Anthropic 24/7 Wall St.
- Alex Karp Is Saying What Every Angry CEO Is Thinking About AI WSJ
- Palantir CEO: “Something Has Gone Completely Wrong” With OpenAI and Anthropic Yahoo Finance
- Opinion | Did We Make the Wrong Bet on Big A.I.? The New York Times
- View / The mistrust of AI labs bubbles over Semafor
Reading Insights
1
3
21 min
vs 22 min read
97%
4,344 → 114 words
Want the full story? Read the original article
Read on 24/7 Wall St.