ChaptersEventsBlog
How is your enterprise using AI Agents? Help us benchmark security and take the survey before November 30 →

Download Publication

The State of Cloud and AI Security 2025
The State of Cloud and AI Security 2025
Who it's for:
  • CISOs
  • Cloud Security Architects
  • IT and Security Managers
  • IAM Professionals
  • Risk and Compliance Officers

The State of Cloud and AI Security 2025

Release Date: 09/09/2025

This global survey report, developed in partnership with Tenable, examines how organizations are adapting security strategies for hybrid, multi-cloud, and AI-driven environments. Drawing on insights from more than 1,000 professionals, it highlights the widening gap between rapid adoption and security readiness.

Today, the majority of organizations operate hybrid environments and use multiple cloud providers. At the same time, AI workloads are moving quickly into production. Over half of organizations are deploying AI and 34% already report AI-related breaches. Despite this, security programs remain reactive by focusing on incidents rather than prevention and relying on basic identity controls.

This report reveals that identity is the biggest cloud risk. It also highlights the growing skills gap and the many ways organizations leave AI systems unprotected. It offers practical recommendations for resetting security strategies around unified visibility, identity governance, and proactive risk management.

Key Takeaways:
  • Over half of organizations (63%) report using more than one cloud provider. Even more (82%) maintain a hybrid infrastructure of some kind.
  • Many organizations (59%) identified insecure identities and risky permissions as the top security risk to their cloud infrastructure. However, many of these same organizations lack the structure or workflows to address these issues at scale.
  • Lack of expertise is the top challenge to securing cloud infrastructure.
  • The most commonly tracked cloud security KPI is security incident frequency and severity. In IAM, the top metric is MFA/SSO adoption rates. Organizations remain focused on surface-level indicators rather than forward-looking measures of performance.
  • More than a third of organizations with AI workloads (34%) have already experienced an AI-related breach.
  • Only 20% of organizations prioritize unified risk assessment, and only 13% focus on tool consolidation. 
Download this Resource

Bookmark
Share
Related resources
Introductory Guidance to AICM
Introductory Guidance to AICM
Capabilities-Based Risk Assessment (CBRA) for AI Systems
Capabilities-Based Risk Assessment (CBRA) for A...
Beyond the Hype: A Benchmark Study of AI Agents in the SOC
Beyond the Hype: A Benchmark Study of AI Agents...
Securing Application-to-Application Traffic with AI/AGI/ML-Powered Virtual Firewalls: A Comprehensive Framework for Multi-Cloud, Hybrid, and On-Premises Environments
Securing Application-to-Application Traffic with AI/AGI/ML-Powered ...
Published: 11/21/2025
Red Teaming Voice AI: Securing the Next Generation of Conversational Systems
Red Teaming Voice AI: Securing the Next Generation of Conversationa...
Published: 11/20/2025
It’s Time to Make Cloud Threat Modeling Continuous
It’s Time to Make Cloud Threat Modeling Continuous
Published: 11/20/2025
Understanding STAR for AI Level 2: A Practical Step Toward AI Security Compliance
Understanding STAR for AI Level 2: A Practical Step Toward AI Secur...
Published: 11/19/2025
Beyond Passwords: FIDO's Authentication Vision for Financial Services in the Cloud Era
Beyond Passwords: FIDO's Authentication Vision for Financial Servic...
December 12 | Virtual
Cloudbytes Webinar Series
Cloudbytes Webinar Series
January 1 | Virtual

Interested in helping develop research with CSA?

Related Certificates & Training