Palo Alto CEO: AI Pricing Must Drop 90% to Drive Adoption
Nikesh Arora warns sky-high token costs are blocking enterprise AI at scale. A 90% price cut is needed.
If you're betting on broad AI adoption, here's a number that should sharpen your focus: 90%. That's how much Palo Alto Networks CEO Nikesh Arora says AI pricing needs to fall before businesses can realistically deploy the technology at scale. That's not a rounding error — that's a structural problem baked into today's AI economics.
Arora's concern centers on token costs, the per-unit pricing model that underpins most large language model usage. Right now those costs are skyrocketing, and for enterprises looking to run AI across thousands of employees or workflows, the math simply doesn't work. You can run a pilot. You can impress the board. But you can't scale what you can't afford.
Read more Palo Alto CEO: AI Pricing Must Drop 90% for Mass Adoption →
This is a real tradeable signal. Any AI infrastructure or hyperscaler play that can credibly compress inference costs — think chip efficiency, model distillation, or competitive commoditization — gets a clearer runway if Arora's thesis proves right. Conversely, companies selling AI access at premium token rates face serious pricing pressure ahead. The CEO of a major cybersecurity firm calling out this dynamic publicly isn't noise. It's a warning shot.
The broader implication is that the AI adoption curve Wall Street is pricing in may be slower and steeper than bulls expect. Enterprise tech rollouts have always lived and died on ROI timelines, and a 90% required cost reduction suggests we're still early — really early — in the scaling phase. Patient capital wins here. Momentum traders should watch the cost-per-token trend the same way they watch oil prices.
Continue reading at US Top News and Analysis
