F-Secure and AMD Silo AI Partner to Secure Agentic AI Workflows

The collaboration centers on F-Secure TrustPath, a security framework that moves away from blanket website trust. Instead, the technology evaluates individual actions, determining whether data should be processed locally on a device or within the cloud. This routing decision is made in real-time based on sensitivity, processing costs, and the specific capabilities of various AI models.

F-Secure CEO Timo Laaksonen emphasizes that the current shift in AI usage creates a critical "Trust Inflection Point." With 84% of consumers questioning the authenticity of online content and only 17% comfortable allowing AI to act on their behalf, the industry faces a significant barrier to adoption. By integrating AMD Silo AI’s compute platform, F-Secure aims to provide granular security that protects the user even when an agent interacts with potentially compromised third-party sites.

Jaakko Vainio, Senior Director at AMD Silo AI, notes that the strategy prioritizes local processing whenever possible to maintain privacy and efficiency. Rather than relying solely on massive, centralized models, the system distributes intelligence to where it is most effective. F-Secure plans to launch the beta version of TrustPath in the fourth quarter of 2026, with a full production rollout scheduled for 2027.