TLDR: A new global survey by NeuralTrust, ‘The State of AI Agent Security 2026,’ indicates that a significant majority (73%) of Chief Information Security Officers (CISOs) are deeply concerned about the security risks posed by AI agents. However, only 30% of organizations possess mature safeguards to address these evolving challenges, highlighting a critical gap between AI adoption and security readiness.
NEW YORK, November 6, 2025 – A recent global survey conducted by NeuralTrust, titled ‘The State of AI Agent Security 2026,’ has unveiled a stark disparity between the rapid adoption of AI agents in enterprises and the preparedness of their security infrastructures. The report, published via PR Newswire, reveals that a substantial 73% of Chief Information Security Officers (CISOs) express ‘very or critical concern’ regarding the security implications of AI agents. Despite this high level of apprehension, a mere 30% of organizations have implemented mature safeguards to protect against these emerging threats.
The survey, which gathered insights from over 160 CISOs and security leaders across North America, Europe, Asia-Pacific, and Latin America between August and September 2025, underscores a critical security gap. According to the findings, one in five organizations (20%) has already fallen victim to at least one AI agent-related security breach. These incidents are primarily attributed to vulnerabilities such as prompt injection and data exposure, indicating that current security measures are often insufficient for autonomous AI systems.
Joan Vendrell, Co-Founder and CEO of NeuralTrust, emphasized the urgency of the situation, stating, “AI agents are now part of enterprise operations, but the security controls protecting them are still catching up. Our findings show that Agentic Security has become one of the most urgent and complex challenges in modern cybersecurity.” The report highlights that many enterprises continue to rely on traditional security controls that were not designed to manage the unique risks introduced by autonomous AI agents.
Also Read:
- Google Cloud’s 2026 Cybersecurity Forecast: A Looming Surge in AI-Powered Cyberattacks and Global Extortion
- Ping Identity Unveils ‘Identity for AI’ to Secure and Empower the Agent Economy
The study also projects a significant increase in AI agent deployment, with one in three enterprises expected to operate more than 500 AI agents by 2028, and over half by 2030. This accelerating shift necessitates a fundamental change in security strategies, as ‘Agentic Security will become a strategic foundation of enterprise trust.’ NeuralTrust’s maturity model further illustrates the current state of readiness, categorizing 46% of enterprises in a ‘Reactive’ tier, 29% in ‘Managed,’ and fewer than 10% achieving ‘Proactive governance.’ This data points to a widespread need for organizations to enhance their security postures to match the pace of AI innovation.
