Survey: Most Security Incidents Involve Identity Attacks

A Permiso Security survey of 512 cybersecurity professionals reveals significant identity-security gaps: despite high confidence in tracking non-human identities, fewer than half have comprehensive visibility, detection of identity-based threats remains slow, and teams rely on multiple tools with heavy manual correlation. Non-human identities—especially AI agents—are rising and often over-permissioned, with most organizations expecting growth and acknowledging access to sensitive data. The findings point to expanding identity sprawl across multi-cloud and multi-IdP environments and forecast increased investment in identity security to improve prevention and real-time visibility.