One in Five Enterprises Has No Real-Time Way to Stop a Runaway AI Agent’s Spending

Quick Facts

• 21% of enterprises rely solely on post-hoc logs to track AI agent spend, with no real-time way to halt runaway executions.
• 85% of enterprises run two or more orchestration platforms, with a mean of 3.1 platforms per organization.
• 64% of companies with revenue above $1 billion reported losses exceeding $1 million tied to AI system failures in 2025.

A new VentureBeat survey of 107 enterprise organizations reveals a sharp gap in AI financial controls: one in five companies has no programmatic way to stop an AI agent before it generates a budget-breaking bill. These organizations learn of the damage only after reviewing logs.

The finding comes from VentureBeat’s July 2026 Pulse Research wave focused on agent orchestration. Respondents are senior and buyer-credible, with 81% serving as recommenders, influencers, or final decision-makers on AI purchases.

How Enterprises Are Trying to Control Agent Spending

Among enterprises that do have controls, approaches vary. Thirty percent rely on native platform controls such as built-in budget caps or throttling. Twenty-five percent have built custom proxy middleware to intercept runaway agents. Another quarter use dynamic routing to shift heavy workloads to lower-cost models.

The remaining 21% use only reactive monitoring. They have no real-time kill switch. Organization size offered little protection: 18% of enterprises with 10,000 or more employees operated under reactive-only controls, compared to 23% of smaller organizations.

Most ‘Agents’ Are Not Agents

The survey also surfaces a deployment reality check. Seventy-one percent of enterprises said a quarter or fewer of their deployed agents can complete multi-step work on their own. Only 10% said true autonomous agents make up the majority of what they run. Gartner has flagged this pattern as “agentwashing” — labeling AI assistants as agents when they lack autonomous capability.

Three percent of respondents said they are still deploying only chatbots. Thirty-five percent said just 1 to 25% of their systems qualify as true orchestration.

Platforms Are Multiplying

Enterprise orchestration stacks have grown fragmented. Eighty-five percent of organizations run two or more orchestration platforms, and 64% run three or more. The average is 3.1 platforms per organization.

Microsoft AI Foundry and Copilot Studio appear in 70% of stacks. OpenAI’s Agents SDK shows up in 68%, and Anthropic’s Claude Platform in 47%. When asked to name a single primary platform, 41% of respondents picked Microsoft and 28% picked Anthropic.

Microsoft CVP of AI Security David Weston acknowledged the governance pressure that comes with fragmentation. “Without a unified control layer, you start to see fragmentation, agents operating in silos, inconsistent governance, and gaps in security,” Weston said. “What customers are asking for is a way to bring order to that complexity.”

The Financial Stakes Are Real

The cost of weak controls is not theoretical. A 2026 Cloud Security Alliance survey found that 65% of enterprises running AI agents experienced at least one agent-related incident in the past 12 months, and 35% of those reported direct financial losses. A separate EY survey found that 64% of companies with revenue above $1 billion reported losses exceeding $1 million tied to AI system failures during 2025.

A 2025 survey found that 85% of companies miss AI cost forecasts by more than 10%, and nearly a quarter underestimate costs by 50% or more. Uber CTO Praveen Neppalli Naga captured the problem plainly: “I’m back to the drawing board, because the budget I thought I would need is blown away already.”

What Comes Next

More than half of respondents (53%) expect their primary control plane to be hybrid by the end of 2026, combining provider-native tools with external orchestration. Only 6% plan to hand control entirely to a provider-managed service. Vendor lock-in, cited by 35% of respondents, is the top reason enterprises resist ceding control to any single model provider.

Switching activity is accelerating. Sixty-eight percent of enterprises plan to adopt, add, or replace orchestration platforms within the next 12 months. Thirty-four percent plan to do so within the quarter. For founders building in this space, the message is clear: budget controls and governance tooling are no longer optional features.

Read more: One in five enterprises can’t stop a runaway AI agent’s spending in real time