Introducing Agentic Evaluation in FAITH
Cisco, Friday, October 23rd, 2026
Cisco introduces Agentic FAITH, extending its Foundation AI evaluation framework to multi-step, tool-using agent tasks.
Cisco says evaluating AI agents requires clear task definitions, consistent execution conditions and a way to examine both process and outcome.
An agent investigating a codebase makes a chain of decisions, such as where to look, which tools to use and when to submit a finding.
Cisco is introducing Agentic FAITH, an extension of Cisco Foundation AI's FAITH evaluation framework.
It adds tasks that require multiple steps and tool use, aimed at cybersecurity and other domains.