How can organizations effectively manage AI usage?
Managing artificial intelligence throughout a large organization without proper usage analytics is equated to issuing every employee a debit card with unlimited spending and no monthly overview. OpenAI has directly addressed this challenge by unveiling updated credit usage analytics and enhanced spending limitations for ChatGPT Enterprise administrators.
The recent enhancements, now available in the global admin console from OpenAI, equip workspace managers with in-depth insights on credit utilization and the ability to set adjustable usage limits on a weekly basis.
What are the key functionalities of these new tools?
This update primarily focuses on two critical functions that were in high demand from IT departments since the introduction of ChatGPT Enterprise in August 2023: improved analytics and effective spending controls.
On the analytics front, the global admin console features advanced Codex analytics. This allows enterprise admins to access user leaderboards, monitor model usage metrics, and gather plugin interaction data.
Regarding spending controls, workspace managers can now establish weekly spending limits for users. These limits are adjustable on a rolling basis, offering choices between notifications and hard restrictions. The implications of this are significant; a notification alerts users that they are nearing their budget, whereas a hard cap outright blocks access when limits are met.
OpenAI’s credit-based pooled pricing structure supports these controls, enabling budgeting at the workspace level rather than enforcing strict per-user limits. For optimal functionality, it is recommended that power users have limits set four to five times higher compared to standard users.
What improvements have been made since ChatGPT Enterprise's launch?
Since its launch in August 2023, the admin console of ChatGPT Enterprise started with basic usage analytics and fundamental administrative features. Over time, enhancements have been ongoing. The User Analytics dashboard saw improvements around May 8, 2025. Further expansions to workspace analytics were rolled out in March 2026. The most recent additions specific to Codex analytics began implementing on May 21, 2026, with additional refinements noted in the June 2026 release notes.
Role-based access control, known as RBAC, plays a crucial role in this framework. With the introduction of granular RBAC for spending controls, OpenAI empowers workspace managers to oversee their team’s spending without needing to seek approval from a global admin for every alteration.
How does this impact organizations that utilize AI?
The pooled credit model with workspace-level budgeting marks a significant shift from the typical inflexible per-user licensing structure. Rather than incurring the same costs for each user irrespective of actual usage, organizations can strategically allocate credits to areas yielding the greatest benefit.
For current users of ChatGPT Enterprise, the immediate takeaway is straightforward: administrators should delve into the new Codex analytics and implement suitable spending caps. The user leaderboard feature is particularly valuable, as it can highlight where AI usage is most concentrated within an organization.
A critical aspect to be mindful of is whether the spending controls are adaptable enough to accommodate organizations with fluctuating workloads. While rolling weekly limits serve steady usage patterns well, companies facing project-based demand spikes may find themselves constrained by caps at inopportune times.