Circles builds AI-native telco experiences on OpenAI API and Codex, lifting ARPU 22% and cutting churn 9%

OpenAI published a new case study on August 3 highlighting how Circles, a telecom services company, is using OpenAI technology to build AI-native telco experiences. Rather than building its own large models, the company builds customer-facing AI capabilities directly on the OpenAI API.

According to OpenAI, Circles has woven personalization into every stage of user operations, creating an AI-native telecom service experience. The case suggests AI is not just a customer-service add-on but a foundation that can move a carrier's core business metrics.

The results are quantified: Circles lifted average revenue per user (ARPU) by 22% and cut customer churn by 9%. For a telecom industry under persistent margin pressure, those numbers carry real weight.

Beyond customer-facing AI, Circles also brought OpenAI Codex into its development workflow, using AI-assisted coding to improve development efficiency and shorten the path from idea to launch.

The significance of the case lies in showing how AI lands in a traditional industry like telecom: personalized, AI-native operations for users on one side, and faster engineering on the other.

For the industry, Circles' results suggest carriers do not need to build large models from scratch; mature APIs can deliver differentiated AI experiences quickly, a route that may become a mainstream path for telco AI transformation.

What to watch next is whether Circles can replicate this AI-native playbook across more markets and operator customers, and whether the OpenAI-telecom partnership deepens further.