Krux
April 4, 2026
97% of Enterprises Expect AI Agent Breach This Year
Published: April 4, 2026 at 12:37 AM
Updated: April 4, 2026 at 12:37 AM
100-word summary
Nearly every enterprise leader surveyed expects a major security incident caused by AI agents within the next 12 months, according to Arkose Labs' survey of 300 Fortune 100-scale companies. Half expect it within six months. The kicker: only 6% of security budgets address AI agent risk, and three-quarters of companies can't even prove when an agent was involved in a breach. The gap reveals a brewing crisis. Companies are handing AI agents credentials and system access without treating them like the insider threats they are. When something goes wrong, most organizations lack the telemetry to know whether a human or an autonomous agent caused it.
What happened
Nearly every enterprise leader surveyed expects a major security incident caused by AI agents within the next 12 months, according to Arkose Labs' survey of 300 Fortune 100-scale companies. Half expect it within six months. The kicker: only 6% of security budgets address AI agent risk, and three-quarters of companies can't even prove when an agent was involved in a breach. The gap reveals a brewing crisis. Companies are handing AI agents credentials and system access without treating them like the insider threats they are.
Why it matters
When something goes wrong, most organizations lack the telemetry to know whether a human or an autonomous agent caused it.