Zscaler 2026 AI Threat Report: 91% Year-over-Year Surge in AI Activity Creates Growing Oversight Gap for Global Enterprises

Core Insights - Rapid AI adoption is creating a significant security gap, necessitating organizations to implement AI security platforms based on Zero Trust principles [1][2][12] AI Adoption Trends - AI usage is expanding across all business functions, with Finance & Insurance accounting for 23% of AI/ML traffic, while Technology and Education sectors saw year-over-year transaction growth of 202% and 184% respectively [4][12] - The number of applications driving AI/ML transactions quadrupled to over 3,400, indicating a rapid increase in complexity and reduced centralized visibility [12] Security Vulnerabilities - 100% of enterprise AI systems tested were found vulnerable to breaches, with critical vulnerabilities identified in an average of 16 minutes, and 90% of systems compromised in under 90 minutes [5][12] - AI/ML activity surged by 91% year-over-year, leading to a total of 18,033 terabytes of data transferred to AI applications, which poses a significant target for cybercriminals [7][9][12] Data Exposure Risks - Unmanaged embedded AI features in enterprise applications are creating critical data exposure risks, as these features often operate without oversight [8][12] - ChatGPT alone was associated with 410 million Data Loss Prevention (DLP) policy violations, highlighting the urgent need for AI governance [10][12] Recommendations for Security - Organizations are advised to adopt an intelligent Zero Trust architecture to mitigate risks associated with AI-driven cyber threats [3][11] - Continuous verification and least-privileged access should be enforced to eliminate attack surfaces, while all traffic, including encrypted data, must be inspected to stop threats in real time [13][11]