Report by Bedrock Security
2025 Enterprise Data Security Confidence Index
Key Findings
64% of organizations aim to strengthen policy enforcement across cloud environments.
57% of security teams find it difficult to enforce policies on training data usage.
83% of Security Engineers/Architects worry most about AI systems understanding data access rights.
More than a third of organisations lack timely visibility into who’s accessing their most sensitive information.
66% of Security Managers/Directors focus on policy enforcement across environments.
20% of CISOs are able to identify over 75% of sensitive data across environments.
5% of Security Managers/Directors have low or no confidence in controlling data used for AI training.
68% of organizations increased focus on infrastructure security while simultaneously taking on new data-centric responsibilities.
86% of professionals reported changes in their security role over the past year as data security duties expand.
55% of Security Managers/Directors added data governance duties for AI training.
Almost 59% of US cybersecurity professionals added new AI data responsibilities in the past year.
Almost 70% of CISOs/CSOs/CTOs have taken on new data discovery responsibilities, specifically for AI initiatives.
83% of CISOs place significantly higher priority on AI data usage governance.
46% of Security Managers/Directors report the lowest confidence in controlling data used for AI training.
62% struggle with different data types (structured, semi-structured, unstructured) as a top barrier to effective data security.
39% of Security Engineers report the highest ability to track sensitive data, able to identify over 75% of sensitive data across environments.
71% of Security Managers/Directors focus on AI governance.
76% of US cybersecurity professionals cite lack of automation requiring too much manual work as a top barrier to effective data security.
Nearly 90% of organisations value a metadata lake approach to solve continuous data discovery and classification challenges.
76% of organisations cannot produce a complete data asset inventory within hours when needed for compliance or security incidents.
65% of organisations need days to accomplish the task of producing a complete data asset inventory.
11% of organisations require weeks or longer to produce a complete data asset inventory.
63% of organisations claim they can identify who accessed specific sensitive data in the last 30 days within 24 hours.
52% of Security Engineers/Architects have new AI data discovery responsibilities.
82% of US cybersecurity professionals blame complex environments with multiple clouds and data stores as a top barrier to effective data security.
72% of CISOs express the most concern about discovering data used in AI initiatives.
Looking ahead, 70% of organisations will focus on AI/ML data usage governance.
Almost 60% of organisations added new AI data responsibilities in the past year.
53% of organizations plan to improve security tools with better data awareness.
68% of organizations will increase infrastructure security focus
59% of security professionals now have new AI data discovery responsibilities.
59% of US cybersecurity professionals see data usage information for non-security needs (cost management, deduplication) as a benefit.
Only 11.5% of US cybersecurity professionals reported no change in their security role in the past year.
53% of security teams lack continuous and up-to-date visibility into their data.
77% of security teams cannot ensure AI systems respect proper data access rights.
79% of security teams struggle to classify sensitive data used in AI/ML systems.
84% of US cybersecurity professionals see a current, accurate data inventory across all systems and data sets as a benefit of a metadata lake.
75% of US cybersecurity professionals report their tools can’t handle current data volumes as a top barrier to effective data security.
58% of organizations want more accurate data classification.
54% of organizations added AI training data governance duties in the past year.
88% of security professionals rated a metadata lake approach as “critical” or “very valuable” to solving their data visibility issues.
Less than half (48%) of organisations express high confidence in controlling sensitive data used for AI/ML training.
97% of CISOs rate metadata lake technology as either “critical” (36%) or “very valuable” (61%) for solving their data visibility and AI governance issues.
64% of security teams have trouble tracking what data feeds their AI systems.
78% of US cybersecurity professionals see better data awareness for security tools as a benefit.
75% of US cybersecurity professionals see enhanced security tool power through data sensitivity awareness as a benefit.
66% of of US cybersecurity professionals cite lack of people and processes for proper analysis as a top barrier to effective data security.