Tenable report highlights escalating cloud risks as AI adoption grows

Tenable, the Exposure Management company, today published its 2025 Cloud Security Risk Report, revealing that cloud workloads supporting artificial intelligence (AI) initiatives are notably more vulnerable than traditional workloads. According to the report, 70% of AI workloads across AWS, Azure and Google Cloud Platform (GCP) contain at least one unremediated critical vulnerability, compared to 50% of non-AI workloads. These findings highlight the increasing security risks organisations face as they integrate AI into core business operations.

Due to their extensive training datasets and complex model development processes, AI workloads are becoming increasingly attractive targets for threat actors. The study revealed that 77% of organisations using Google’s Vertex AI Workbench had at least one notebook instance configured with an overprivileged default service account. This misconfiguration presents a significant risk, potentially enabling privilege escalation and lateral movement within cloud environments. As AI adoption accelerates in India, the findings underscore the importance of integrating security measures early in the AI development lifecycle.

The report also highlights progress in managing broader cloud security risks. So-called "toxic cloud trilogies", workloads that are simultaneously publicly exposed, critically vulnerable, and excessively privileged, were found in 29% of organisations surveyed, representing a nine-percentage-point decrease from 2024. Tenable attributes this improvement to more effective risk prioritisation and wider adoption of cloud-native security tools. However, the researchers caution that even a single toxic workload can offer attackers direct access to sensitive data.

Identity management continues to be a foundational element of secure cloud environments. The report found that 83% of AWS users have configured at least one identity provider (IdP), which is considered a best practice for securing both human and service identities. Despite this, identity-related risks persist. Credential abuse remains the most common initial access vector, implicated in 22% of breaches. The findings emphasise that merely adopting IdPs is insufficient without robust enforcement of multi-factor authentication and least-privilege access policies.

With India moving towards AI and cloud regulation under the forthcoming Digital India Act, organisations are urged not to delay in securing their cloud AI workloads. The rapid pace of innovation in this space presents significant security challenges, and without a comprehensive cloud security strategy, the risk of cyberattacks remains high.

“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 findings are based on telemetry data from workloads across various public cloud and enterprise environments, collected and analysed by the Tenable Cloud Research team between October 2024 and March 2025.