The Accountability Gap: The State of AI Agent Security and Governance 2026

How security leaders are building visibility, attribution, and accountability for autonomous systems

Within five years, 93% of cybersecurity leaders expect AI agents to access enterprise systems more often than humans. As agents gain authority, governance is moving from a policy question to an operational one: which agents are running, what they did, and who owns the outcome.

In this joint research report from SecurityHQ and Datadog, 204 cybersecurity leaders share how they monitor, govern, and investigate agent activity across clouds, platforms, and third parties. Read the report to explore:

• The six capabilities that make agent governance work in practice, from visibility and attribution to reconstruction and clear accountability
• Why real-time monitoring and auditability build more trust in agents than explainability
• Where leaders still require mandatory human review of agent decisions
• How healthcare, manufacturing, financial services, technology, and retail compare on agent readiness

Read the report.