Bottom Line: Security incidents in the AI environment are widespread, but lack of central transparency and governance prevent organizations from fully controlling risks.
A DigiCert study shows that AI security risks have already become a practical reality — 78 percent of organizations have experienced incidents or identified vulnerabilities. Despite rapid technology rollout, many lack transparency and governance.
78 percent of surveyed organizations report AI-related security incidents or vulnerabilities. In parallel, deployment velocity is growing: three out of four companies have introduced at least four new AI-enabled systems in the past six months. This discrepancy between growth and security readiness characterizes the current state in enterprise IT departments.
The central challenge lies in missing visibility. According to the study, nearly one in two organizations lacks central transparency regarding deployed AI systems and their activities. Concurrently, 45 percent of organizations cannot fully trace which models and data sources their AI decisions are based on. This lack of traceability complicates incident response and jeopardizes compliance requirement adherence.
Governance remains a theory-practice problem: while approximately 90 percent of companies discuss AI governance at leadership level, only about half have established a formal governance program. Still, a countertrend emerges: over 50 percent are now investing intentionally in AI security through dedicated budgets. 86 percent of organizations have implemented processes to revoke access rights from compromised AI systems — a containment mechanism for incidents. Additionally, many enterprises assign digital identities to their AI agents to make activities traceable.
On the regulatory front, momentum is building: 90 percent of respondents have already assessed their liability risks and are preparing for stricter regulatory requirements. CISOs must anticipate that both internal governance requirements and external regulations will tighten in the coming months — while simultaneously the number of productive AI systems in their organizations grows.
Source: www.it-daily.net · Published July 10, 2026
Lumi AI News — AI-assisted curation according to Art. 50 EU AI Act. Paraphrase and classification through Lumi News Pipeline v1.7.3.