CYBERSECURITY RESEARCH
Investigate Exposures. Connect the Evidence. Inform Defense.
Understanding a security finding requires more than a severity label. It requires examining the affected assets, the conditions that make exploitation possible, and the evidence supporting the conclusion.
Kodaowl’s research focus connects AI security with continuous threat exposure management—exploring how weaknesses emerge, interact, and can be addressed across modern technology environments.
AI Agents & MCP Security
Investigate how agent permissions, tool connections, delegation, and untrusted inputs can introduce security exposures.
Vulnerabilities & Exploitability
Examine vulnerability disclosures, exploitation prerequisites, and threat intelligence to distinguish potential weaknesses from relevant exposures.
Exposure Paths & Dependencies
Explore how relationships among applications, identities, cloud resources, AI systems, and software components can connect individual weaknesses into broader risk.
Remediation & Validation
Examine corrective actions, validation methods, and the evidence needed to assess whether an exposure has been addressed.
Research With Clear Boundaries
Our approach emphasizes traceable sources, explicit assumptions, and clear distinctions between observed behavior, inferred risk, and untested scenarios. Research findings should explain their scope and limitations alongside their implications.
From Research to Product Learning
Research informs the development of Koda AI Security and Koda CTEM—helping shape exposure analysis, prioritization, and validation methods. Research topics and experimental findings do not necessarily represent available product capabilities.
Explore Research Publications
Technical analyses, research notes, and practical findings will be published here as they become available.
Decide with Evidence. Act with Confidence.