AI coding agents could soon cost more than the developers using them


Source URL: www.theregister.com/ai-and-ml…
The Register reports Gartner’s warning that AI coding-agent economics are becoming difficult for software engineering teams to govern as vendors move from seat-based licensing to consumption-based pricing. Gartner cited examples of monthly AI coding bills rising from tens or hundreds of dollars to $2,000–$5,000 per developer, with extreme cases reaching $20,000 in token charges, while engineering leaders often lack transparency into how token consumption is calculated or controlled. The governance issue is not simply cost; it is the combination of opaque consumption, weak cost-optimization tooling, uncertain productivity correlation and the need for stronger practices such as context engineering, model routing and differentiated use of frontier models for only the highest-value tasks.