Addressing mounting concerns over escalating costs of corporate artificial intelligence (AI), OpenAI on Thursday launched a suite of usage analytics and enhanced spending controls for its ChatGPT Enterprise platform.
The new administrative tools are designed to provide corporate clients with unprecedented visibility in their AI investments and finer budgetary control over deployments.
The update introduces an overhaul to OpenAI’s Global Admin Console, consolidating ChatGPT and Codex credit use into a single, centralized dashboard. Company administrators can now view granular breakdowns of AI consumption filtered by individual users, specific business units, and distinct AI models. This allows organizations to track adoption trends over time, pinpoint power users, and identify shifting operational patterns.
Crucially, the rollout acknowledges growing demand for cost governance as enterprise-wide AI consumption swells. Account managers can now establish baseline credit limits across an entire workspace, set customized budgets for different departments, and issue individual overrides for employees requiring high-capacity access. On the user end, employees can monitor their own consumption metrics and submit justified requests for additional credits directly through the system.
The feature rollout arrives as corporate enthusiasm for generative AI undergoes a significant shift from rapid adoption toward strict financial oversight. Industry analysts note that while initial deployments were fueled by experimentation, organizations are now grappling with the unpredictable costs of usage-based AI economics, which are tied to volatile metrics like token volume, large language model (LLM) requests, and GPU hours.
“Enterprises are clearly shifting from adoption-led enthusiasm to cost and value governance,” Forrester analyst Biswajeet Mahapatra said. “AI is no longer an adoption problem but a measurement and credibility problem.”
While analysts viewed OpenAI’s new dashboard as a foundational step toward operational control, they cautioned that tracking expenditure is only half the battle. Present tools remain unable to directly map AI costs to tangible business value, leaving productivity gains largely fragmented and difficult to quantify on a balance sheet.
“Token consumption alone is insufficient because it measures activity rather than impact,” Mahapatra added, suggesting that corporations eventually marry these consumption metrics with traditional indicators like revenue growth and risk mitigation.
The urgency for robust AI financial operations (FinOps) is expected to intensify rapidly.
Anushree Verma, senior director analyst at Gartner, highlighted that inconsistent vendor pricing structures and fragmented consumption models currently make long-term forecasting notoriously difficult. Verma projected that by 2028, an average global Fortune 500 enterprise will deploy more than 150,000 autonomous AI agents — a massive leap from fewer than 15 in 2025. This imminent agent sprawl increases the risk of interconnected system misconfigurations, which can cause real-time costs to spike exponentially.
By equipping administrators with immediate budgeting controls, OpenAI aims to provide a buffer against these compounding IT complexities. The updated analytics and spend management features became active for all ChatGPT Enterprise clients starting last week.
