Tokens or humans? The new corporate trade-off Artificial intelligence is proving more expensive than anticipated, leading major U.S. companies to confront a difficult trade-off between AI token costs and human employees. Enterprise AI CEOs report that company AI budgets are being exhausted within one to two months, an unsustainable path due to escalating token prices per model release. This situation prompts a comparison where technology costs are now equivalent to personnel costs, a historically unprecedented scenario.

• Enterprise AI budgets are being exhausted rapidly, with annual budgets depleted in one to two months.
• The cost of AI, particularly per token, has increased with each new model release.
• Companies are now facing a trade-off between investing in AI technology and maintaining human staff.
• AI’s inefficiency means its current value generation lags behind business costs.
• A significant portion of enterprise AI usage (95%) relies on expensive frontier models for tasks that could be handled by cheaper alternatives.
• Optimizing AI spend involves routing tasks to the most cost-effective model, offering potential for substantial savings.