Businesses are rolling out AI agents across their networks much faster than they can implement monitoring procedures and security controls, leaving organizations exposed to security threats, hallucinations and other common issues associated with the technology.

Published by Cisco on October 20, the2026 Splunk State of Observability warned that while many businesses are keen to explore the potential benefits of using AI in their ecosystem, many do so without a strategy in place to ensure it is being used safely.

According to the 2050 ITOps and engineering respondents, 38% said their organization does not yet have clearly defined ownership and accountability for monitoring AI.

Much of this confusion stems from how responsibility for deployment and monitoring of AI agents cuts across traditional boundaries within the IT department such as software engineering, infrastructure management, web development and information security.

Read more: Agentic AI will Supercharge Cyber Threats. But Not the Way You Think

This fragmentation has made it difficult to monitor the use of AI within the organization and 42% of those surveyed said that their main challenge in observing AI is that a different team is responsible for it.

“When agent observability is everyone’s job, it becomes nobody’s job,” said Mark Woods, field CTO and chief technical advisor for EMEA at Splunk.

Organizations Struggle to Manage AI Agents at Scale

Many IT and security teams are also struggling to keep up the potentially unintended consequences of swift rollout of AI tools within the enterprise. For example, while 85% of respondents said that having visibility into AI hallucinations is important, only 58% said that their organization can effectively monitor for potential AI hallucinations.

For many organizations, identifying when an AI agent is making or acting upon an error is difficult and they might not have a means of preventing the action from happening.

This could lead to negative consequences for the business. For example, a malicious attacker could take advantage of an AI agent to trick it into exposing sensitive information. If an organization had appropriate monitoring tools in place, this kind of activity could be prevented.

"Managing agents at scale requires reimagining both security and observability. Have our agents been compromised? Are they doing their job well? In the world of agentic AI, these questions are inseparable, and they're moving at machine speed,” said Jeetu Patel, president and CPO at Cisco.

“This report shows that observability and security data are converging, and that teams need to collaborate more than ever. Common data platforms that help us detect and respond to issues in real-time are the key to deploying agents we can trust at scale,” he added.

The report also noted that when AI agents are deployed correctly, they can provide a significant boost to IT and security teams.

Of those surveyed, 86% said that the deployment of AI had improved the speed at which they are able to detect IT or security issues and 36% classed that improvement as “significant”.

However, while the majority of those surveyed said that agentic AI had made them more efficient at work, some disagreed: 14% suggested that the deployment of AI agents had made issues harder to detect and resolve. One of the main issues cited was the extra time it took to double check if what the agent was saying was correct.