Cisco Foundation AI is getting new agentic security tools to help companies secure autonomous AI systems. These include Foundation-sec-8B-Reasoning, a cybersecurity reasoning model; an adaptive AI search framework; and the open-source threat-hunting assistant PEAK. The tools offer transparency and keep people involved in the decision-making process.
Foundation-sec-8B-Reasoning is an open-weight reasoning model for cybersecurity use cases. The model is optimized for multistep security analysis, including threat modeling, attack path analysis, and incident investigation.
The model produces explicit reasoning traces that allow analysts to understand how conclusions are reached. This supports validation, trust, and regulatory requirements in high-impact security environments. The focus on transparency is crucial. After all, organizations must be able to explain how AI systems arrive at security decisions.
Adaptive search framework and threat hunting assistant
Cisco is also introducing the Adaptive AI Search Framework, a reasoning-driven information-retrieval system that goes beyond static, query-based searches. The framework enables AI models to iteratively refine their search strategies as new information emerges, much like a human security expert would.
The third release is PEAK Threat Hunting Assistant, an open-source agentic AI assistant that automates threat-hunting preparation. It allows AI agents to collaborate to investigate threat actors and techniques, analyze internal security data, and generate customized, step-by-step threat hunt plans. Cisco Foundation AI emphasizes that PEAK keeps humans in the loop, allowing security teams to maintain control over decisions, models, and data access.
Tip: Cisco adds more security for AI apps to Cisco Security Cloud