The invisible cost of tokens: why the AI bill keeps rising

▤ Summary

The article explains why AI costs are rising even as model prices fall. It shows how token-based usage, especially in agentic workflows, can drive bills far higher than companies expect. It highlights examples from GitHub, Uber, and Cisco to show how enterprises are reacting with usage limits and tighter governance. The piece also points to a growing need for cost controls, vendor contract renegotiation, and better measurement of AI tasks by business value rather than raw token volume. Overall, it frames AI spend management as an emerging operational priority for software buyers.

▦ Linked Companies