Enterprises Adopt Multiple AI Orchestration Platforms Amid Trust Concerns
A majority of enterprises are now using multiple AI orchestration platforms due to a lack of full trust in single vendors for security and permissions, leading to more complex solutions.
Enterprises are moving away from relying on single AI orchestration platforms, with the average company now using three simultaneously. This shift is driven by a lack of complete trust in any single vendor's security and permissioning capabilities, according to VB Pulse data.
The data indicates that 85% of enterprises utilize at least two orchestration tools, and 64% use three. Only 15% of companies are committed to a single platform. While Microsoft currently leads in primary usage, Anthropic is a strong contender for future adoption.
Key factors influencing enterprise decisions include flexibility (29%), security and permissions (17%), and execution reliability (15%). A significant challenge remains the uncontrolled spending by AI agents, with one in five enterprises unable to stop runaway spending in real time.
To manage costs, companies are employing strategies such as native platform controls (30%) or custom gateway software (25%). Additionally, 25% use dynamic routing to leverage less expensive models, while 21% rely solely on post-hoc monitoring without real-time kill switches.
The trend towards hybrid solutions is expected to continue. Over half of respondents (53%) anticipate a hybrid control plane by the end of 2026, signaling a strategic move towards integrating multiple platforms, models, and agents.