A survey of mainly large APJ organizations has, besides AI threats, unearthed gaps in executive awareness and detection of insider risks.
Based on a survey of 1,010 cybersecurity professionals* across multiple industries and regions in June and July 2025 about how AI and the analytics gap are fueling insider risks in cybersecurity, a cybersecurity firm has shared its data findings on insider threats fueled by AI and analytics gaps with the media.
First, 69% of respondents in the Asia Pacific and Japan (APJ) region had indicated an expectation that insider threats would increase in the next 12 months — the region with the highest anticipation of in insider threats in the survey. Additionally, 60% had cited a measurable increase in insider incidents during the past year.
Second, 75% of APJ respondents had indicated that AI was increasing the effectiveness of insider attacks, with AI-enhanced phishing and social engineering identified by 31% as the top insider threat vector, followed by privilege misuse (18%) and data exfiltration (17%).
Other findings
Third, 53% of APJ respondents viewed insider threats — whether malicious or compromised insiders — as a risk equal or greater to that from external threat actors. Also:
- Unauthorized use of generative AI tools by employees was reported by 64% of APJ respondents, with 12% identifying this as their top insider threat concern.
- 82% of APJ respondents cited having insider threat programs, and 37% cited using user and entity behavior analytics for detecting abnormal activity.
- 94% of APJ respondents reported some use of AI, with over half of the global survey sample indicated their belief that these tools were fully deployed despite many still being in pilot or evaluation phases.
- A disconnect was noted between security teams and executive leadership, with 74% of respondents globally citing that executive leadership underestimated insider threats.
According to Steve Wilson, Chief AI and Product Officer, Exabeam, the firm releasing its data findings: “Insiders aren’t just people anymore. They’re AI agents logging in with valid credentials, spoofing trusted voices, and making moves at machine speed. The question isn’t just who has access — it’s whether you can spot when that access is being abused.”
*including analysts, security team leads, and executive decision-makers employed in technology, financial services, manufacturing, healthcare, retail, and government, with a significant proportion from large enterprises employing 500 or more staff across parts of the Asia Pacific and Japan (APJ), North America, Europe, and the Middle East regions. No breakdown of respondent profiles by region is disclosed in the full survey report.