New credit usage analytics and spend controls for ChatGPT Enterprise

OpenAI is addressing a growing tension for enterprises: as AI usage scales across organizations, visibility into consumption and cost control has lagged behind deployment. Without granular tracking, companies struggle to distinguish productive adoption from wasteful usage, making it hard to defend AI spend or scale responsibly. The article frames this as a governance gap that threatens enterprise confidence in AI investments.

The concrete response lands in the Global Admin Console, which now consolidates ChatGPT and Codex credit usage into a unified view. Admins can track trends over time, identify top users, and break down spend by user, product, or model. On the control side, OpenAI extends earlier custom role limits with workspace-wide default caps, group-specific limits, and individual overrides — plus end-user visibility into their own budget. Employees can request limit increases with context, letting admins approve capacity for power users without raising everyone’s ceiling.

The takeaway for builders and ops teams: these are not just dashboard features but infrastructure for scaling AI governance. The unified Cost API lets organizations pull data into their own systems for deeper analysis, which signals that OpenAI expects enterprises to integrate these controls into existing FinOps or procurement workflows. The real move here is making AI spend auditable and discretionary — a prerequisite for treating AI as a managed business investment rather than an experiment.