AI pricing began as a nightmare – now it is even worse

▤ Summary

The article examines how hard it is for enterprises to price AI in a way that ties cost to business value. It argues that token-based pricing and other usage models fit vendors better than customers because they measure consumption, not outcomes. The piece also warns that pure ROI-based pricing can push AI systems toward harmful optimization and unintended behavior. It recommends governance through an AI committee, clear benefit and risk quantification, and executive accountability tied to incentives before any pricing agreement is set.

▦ Linked Companies

SAP

$1B+