John Brown
Member
Thales has officially launched AI Security Fabric to enhance AI runtime security, introducing a new defensive framework that protects Agentic AI, large language model (LLM)-powered applications, enterprise data, and identities from evolving AI-related threats. Designed to secure AI workflows in real time across cloud, on-premises, and hybrid environments, this breakthrough platform helps organizations innovate with confidence while mitigating risk and maintaining compliance.
What the AI Security Fabric Does Runtime Protection for AI Ecosystems
AI is rapidly reshaping business operations, but it also brings new security vulnerabilities unique to autonomous and generative systems. Thales' AI Security Fabric addresses these risks by providing runtime capabilities that monitor and protect AI activity as it happens, helping organizations to:- Mitigate AI-specific threats: The platform defends against prompt injection, model manipulation, sensitive data leakage, jailbreaking, and other attacks that target LLM-based applications.
- Safeguard enterprise assets end to end: By integrating controlled dataset access, encryption, and key management, the solution protects sensitive information and identity interactions throughout AI operations.
- Support diverse environments: True runtime protection can be deployed in cloud-native, on-premises or hybrid architectures helping businesses secure AI interactions regardless of where models run.
Why Runtime Security Matters for AI Adoption
As enterprises increase AI adoption across business functions, the attack surface grows, and so do compliance challenges. According to industry research, a large majority of organizations now use AI in at least one function, and many are allocating budget to AI-specific security tools to protect sensitive operations.Thales' approach shifts security from design time to runtime, meaning the platform continuously observes and defends AI behavior as models interact with users and data. This is important because AI systems can behave unpredictably once deployed, and static pre-deployment checks are often insufficient to catch runtime attacks
Future Expansion Deeper Runtime Capabilities in 2026
Thales plans to expand the AI Security Fabric in 2026 with enhancements that further fortify runtime protections, including:- Data leakage prevention at scale — stronger safeguards to detect and block sensitive data exposure.
- Model Context Protocol (MCP) security gateway — monitoring and filtering AI model access to external systems.
- End-to-end runtime access control — unified access policies across AI agents, data, and user interactions.
Expert Perspective
Sebastien Cano, Senior Vice President of Thales' Cyber Security Products Business, emphasizes that AI introduces risks that traditional security solutions aren't equipped to handle. The AI Security Fabric provides organizations specialized tools to protect AI applications while simplifying operational complexity making safe, compliant scaling of AI easier for enterprises of all sizes.Surgery
Thales' AI Security Fabric represents a significant step forward in AI runtime security, offering enterprises the ability to protect GenAI and Agentic AI systems in real time. With its focus on threat mitigation, data protection, and flexibility across environments, the platform helps organizations unlock AI's potential without sacrificing security or compliance, an essential capability as AI becomes foundational to modern business operations.SOC News provides the latest updates, insights, and trends in cybersecurity and security operations.
Read related news - https://soc-news.com/amentum-secures-dtra-prime-contract-for-defense-projects/