GUEST RESEARCH: Cloud AI workloads riskier than traditional with 70% containing critical vulnerabilities compared to 50% in non-AI workloads.
Tenable, the Exposure Management company, has released its 2025 Cloud Security Risk Report, revealing that cloud workloads supporting artificial intelligence (AI) initiatives are more vulnerable than traditional workloads. The report found that 70 per cent of AI workloads across AWS, Azure and GCP contain at least one unremediated critical vulnerability, compared to 50 per cent of non-AI workloads, highlighting the mounting risk as organisations embed AI into their business operations.
AI workloads, with their vast training datasets and model development processes, are an increasingly attractive target for threat actors. The study found that 77 per cent of organisations using Google’s Vertex AI Workbench had at least one notebook instance configured with an overprivileged default service account, a misconfiguration that could open a gateway for privilege escalation and lateral movement across cloud environments. As AI adoption accelerates in Australia, the findings underscore the need for organisations to embed security earlier into AI development lifecycles.
Tenable’s research also shows broader progress in cloud risk management. Toxic cloud trilogies, workloads that are publicly exposed, critically vulnerable, and highly privileged, fell to 29 per cent of organisations surveyed, a nine-point improvement from 2024. Tenable’s researchers attribute the nine-point decline to sharper risk-prioritisation practices and wider use of cloud-native security tooling, yet warn that even a single trilogy provides attackers with a fast lane to sensitive data.
Identity remains the foundation of a secure cloud environment. The report finds that 83 per cent of AWS users have configured at least one identity provider (IdP), a best practice for securing human and service identities. Yet, the presence of identity-based risks persists. Credential abuse remains the most common initial access vector, implicated in 22 per cent of breaches, underscoring that simply adopting IdPs is not enough without strong enforcement of multi-factor authentication and least-privilege principles.
“Organisations have made real strides in tackling toxic cloud risks, but the rise of AI workloads introduces a fresh wave of complexity,” said Ari Eitan, Director of Cloud Security Research at Tenable. “AI’s data-intensive nature, combined with persistent misconfigurations and vulnerabilities, demands a new level of diligence. Exposure management gives security teams the context they need to protect what matters most, including the crown jewels hidden inside AI environments.”
The report reflects findings by the Tenable Cloud Research team based on telemetry from workloads across diverse public cloud and enterprise environments, analysed from October 2024 through March 2025. To download the report today, please visit: https://www.tenable.com/cyber-exposure/tenable-cloud-security-risk-report-2025?utm_medium=referral&utm_source=na&utm_campaign=cmpn-00033589&utm_content=apj_bylines