Gurucul Survey Finds Rising Insider Incidents and Gaps in AI Risk Controls
Summary
Gurucul and Cybersecurity Insiders surveyed 725 IT and security professionals: 90% reported insider incidents, 54% reported confirmed or suspected AI-related incidents, and only 20% expressed confidence in containing AI-related incidents before major damage.
Key points
- The report draws on an online survey of 725 IT and cybersecurity professionals conducted in late 2025; comparisons use a 2024 study of 413 professionals.
- Among respondents, 90% reported at least one insider incident in the previous year, 56% reported six or more, and 55% said incidents had become more frequent.
- Negligent insiders were a concern for 74% of respondents, ahead of compromised accounts at 65% and malicious insiders at 59%.
- AI adoption increased perceived insider-risk exposure for 94% of respondents. Separately, 23% reported confirmed AI-related insider incidents and 31% reported suspected incidents.
- Detection accuracy challenged 66% of respondents, while 58% cited tool and data fragmentation; 34% used five or more insider-risk tools.
- Only 20% expressed confidence in detecting and containing AI-related insider incidents before significant damage. Among organizations using AI in insider-risk workflows, 57% reported successful alert-triage automation versus 26% for response and containment.
- The report recommends integrating identity and behavioral visibility, governing AI access as insider access, and connecting detection to automated investigation and containment.
Article Details
- Publisher
- Cybersecurity Insiders and Gurucul
- Report Period
- Survey conducted in late 2025; incident questions cover the preceding 12 months, with comparisons to the 2024 study.
- Scope
- Enterprise insider risk management, AI-related exposure, detection and response capabilities, tooling, and governance.
- Sample Size
- 725 IT and cybersecurity professionals surveyed online; reported margin of error approximately ±3.6 percentage points at 95% confidence. The comparison study surveyed 413 professionals.
- Key Statistics
- 90% of respondents reported at least one insider incident at their organization in the preceding 12 months; 56% reported six or more.
- 94% reported that AI adoption increases insider risk exposure; 74% described the increase as moderate or significant.
- 23% reported confirmed AI-related insider incidents in the preceding 12 months and 31% reported suspected incidents; a separate 25% believed incidents were likely occurring undetected.
- Only 20% were confident or very confident their organization could detect and contain an AI-related insider incident before significant damage occurred.
- Among organizations using AI in insider risk workflows, reported automation success was 57% for alert triage and risk scoring, 39% for investigation and case management, and 26% for incident response and containment.
- Recommendations
- Integrate identity, access, and behavioral data to establish unified visibility before adding more security tools.
- Govern AI as an insider with delegated authority, extending visibility and access controls to AI agents and machine identities.
- Use AI automation to scale detection and response while retaining human expertise for contextual judgment and high-impact decisions.
- Connect detection to investigation and containment through integrated data, automation, and response orchestration.