Survey reveals 20% of enterprises lack real-time AI cost controls

A recent survey of 107 enterprises reveals that organizations are increasingly adopting multiple orchestration platforms to manage their AI agents. On average, these companies utilize three different platforms, prioritizing flexibility over allegiance to any single provider. Microsoft is currently the most widely used platform, while Anthropic is gaining traction for future consideration.

This trend is driven by the need for businesses to maintain control over their AI systems in a rapidly evolving landscape. As companies integrate AI into their operations, they face significant challenges in ensuring that these systems operate effectively and securely. Enterprises are designing their AI control structures to be hybrid, combining leading AI providers with independent technologies. This approach mitigates risks associated with provider lock-in and enhances security against potential vulnerabilities tied to specific platforms.

However, the research highlights a critical gap in real-time monitoring and control. Approximately 20% of surveyed enterprises lack a mechanism to halt a runaway AI agent before incurring unexpected costs. This inability to manage expenses in real-time poses significant financial risks, as companies increasingly rely on AI for various business functions. Without effective oversight, organizations could face spiraling costs as AI agents operate autonomously.

Looking ahead, the findings underscore the importance of developing robust governance and oversight mechanisms in enterprise AI deployments. As companies continue to embrace AI, the need for effective orchestration and cost management will become even more vital. Establishing clear guidelines and controls can help organizations harness the benefits of AI while minimizing the associated risks.