2.6K
Baltimore, MD, December 2nd, 2025, CyberNewsWire
The 2025 State of AI Data Security Report unveils a concerning trend in enterprise security: while AI adoption is widespread, oversight remains inadequate. A staggering 83% of organizations incorporate AI into their daily operations, yet only 13% claim to have a comprehensive understanding of how these systems manage sensitive data.
Commissioned by Cybersecurity Insiders in collaboration with Cyera Research Labs, the study surveyed 921 cybersecurity and IT professionals across various industries and company sizes.
The data highlights AI’s evolution into an unregulated entity – a non-human user with accelerated processing capabilities, extensive access privileges, and continuous operation. However, most organizations rely on human-centric identity frameworks that struggle to keep pace with AI’s rapid actions. Consequently, two-thirds of respondents have observed AI tools overreaching their data access boundaries, with 23% admitting to lacking controls over AI prompts and outputs.
Autonomous AI agents emerge as the most vulnerable frontier, with 76% of participants citing difficulty in securing these systems. Additionally, 57% lack the capacity to swiftly block risky AI behaviors in real-time. Visibility into AI usage remains scarce, as nearly half of the respondents report no insight into AI activities, while a third possess only minimal understanding – leaving many enterprises uncertain about AI’s operations and data interactions.
Furthermore, governance structures struggle to keep pace with AI adoption rates. Merely 7% of organizations boast a dedicated AI governance team, and only 11% feel adequately equipped to comply with evolving regulatory mandates, underscoring the widening readiness gaps.
The report advocates for a shift towards data-centric AI oversight, emphasizing continuous monitoring of AI usage, real-time supervision of prompts and outputs, and identity policies that treat AI as a distinct entity with limited access based on data sensitivity.
“AI is no longer just a tool – it functions as a new identity within organizations, one that operates incessantly and often disregards boundaries,” remarked Holger Schulze of Cybersecurity Insiders. “Without robust governance and visibility, enterprises risk finding their data in unauthorized locations.”
The report warns: “You cannot secure an AI agent you cannot identify, and you cannot govern what you cannot see.”
For the complete 2025 State of AI Data Security Report, visit: https://www.cybersecurity-insiders.com/portfolio/2025-state-of-ai-data-security-report-cyera/
Media Contact: [email protected]
About Cybersecurity Insiders
Cybersecurity Insiders offers strategic insights for security leaders, drawing on over a decade of independent research and trusted by a global network of 600,000 cybersecurity professionals. The platform translates evolving market trends into actionable recommendations that empower CISOs to fortify their security programs, make informed technology choices, and anticipate emerging threats.
By bridging the gap between practitioners and innovators, Cybersecurity Insiders equips CISOs with the clarity needed to navigate a complex industry landscape, while helping solution providers align with real-world priorities. This alignment is achieved through evidence-based research, strategic CISO guides, impartial product evaluations, data-driven message validation, and peer-validated accolades such as the Cybersecurity Excellence Awards and AI Leader Awards.
Learn more at: https://cybersecurity-insiders.com
Contact
Founder
Holger Schulze
Cybersecurity Insiders
[email protected]
