Cisco Foundation AI, the research and engineering division of Cisco Systems Inc., has unveiled a new suite of agentic security tools aimed at enhancing the security of AI systems as they become increasingly autonomous and integrated into business operations. This announcement comes as enterprises face mounting challenges in safeguarding AI-driven environments that span multiple platforms, including cloud services and internal systems.
The new tools are designed to assist AI systems in navigating complex security issues by enabling them to dynamically retrieve evidence from fragmented data sources and execute security workflows while ensuring transparency and human oversight. Cisco’s Foundation AI team contends that traditional security methods are struggling to keep pace with the complexities and rapid decision-making capabilities of modern AI workflows, particularly as these systems begin to operate independently.
At the forefront of these developments is the launch of Foundation-sec-8B-Reasoning, an open-weight reasoning model specifically optimized for cybersecurity applications. Unlike general-purpose language models, this model is tailored for multistep security analysis, encompassing threat modeling, attack path analysis, and incident investigation. It generates explicit reasoning traces that help security analysts understand the rationale behind conclusions, thus bolstering validation processes and compliance with regulatory requirements in high-stakes security scenarios.
Cisco also introduced the Adaptive AI Search Framework, a next-generation information retrieval system that moves beyond static search queries. This framework empowers AI models to iteratively refine their search strategies as new data emerges, mirroring the adaptive thinking of human security experts. Its design aims to enhance threat intelligence analysis and incident response by allowing AI systems to adjust their investigative paths when faced with incomplete or noisy data.
The third significant release is the PEAK Threat Hunting Assistant, an open-source agentic AI assistant dedicated to automating the preparation for threat hunting activities. PEAK utilizes cooperative AI agents to investigate threat actors and techniques, analyze internal security data, and create customized, step-by-step threat hunt plans. Cisco emphasizes that this tool ensures human oversight, allowing security teams to maintain control over decision-making, model interpretations, and data access.
“Collectively, the reasoning model, AI search framework, and PEAK reflect how Cisco Foundation AI is delivering disproportionate impact by addressing foundational challenges at the intersection of AI and security,” said Yaron Singer, vice president of AI and security at Cisco Foundation AI, in a blog post. He highlighted that Cisco’s approach emphasizes open, security-native foundations, enterprise deployability, and architectural rigor.
The introduction of these tools underscores a growing recognition of the urgent need for robust security measures tailored specifically for AI systems. As organizations increasingly embed AI technologies into their operations, the potential for vulnerabilities escalates, necessitating innovative solutions that can keep pace with rapid advancements in AI capabilities.
With enterprises navigating an ever-evolving threat landscape, the security tools from Cisco Foundation AI may prove vital in fortifying defenses while enabling the autonomous capabilities of AI systems. As these technologies become more commonplace, the focus on blending security with AI innovation will likely shape the future of enterprise cybersecurity.
See also
SpacemiT Launches K3 AI CPU with 60 TOPS to Advance Open-Source RISC-V Ecosystem
Tesseract Launches Site Manager and PRISM Vision Badge for Job Site Clarity
Affordable Android Smartwatches That Offer Great Value and Features
Russia”s AIDOL Robot Stumbles During Debut in Moscow
AI Technology Revolutionizes Meat Processing at Cargill Slaughterhouse
