Investigating, analyzing and mitigating emerging threats targeting AI / ML to build a world-class AI threat research capability, delivering actionable intelligence, advancing the state of AI security research, and helping secure Cisco’s AI-driven products and services.
Requirements
- 5+ years of experience in cybersecurity threat intelligence, adversarial research, red teaming, or offensive security, with exposure to AI/ML systems preferred.
- Strong foundational knowledge of AI/ML technologies, particularly LLMs, generative models, and agentic frameworks (e.g., LangChain, Hugging Face Agents).
- Proficiency in Python or similar scripting languages for research tasks, threat simulations, and automation.
- Experience conducting threat research, writing technical analyses, and contributing to security investigations or adversarial projects.
- Ability to clearly document and present research findings to diverse technical audiences.
- Contributions to security research projects (whitepapers, blogs, open-source tools, or community participation).
- Familiarity with AI-specific threat frameworks such as MITRE ATLAS, ATT&CK for AI, or equivalent.
Responsibilities
- Conduct original research into AI/ML-specific threats, including prompt injection, data poisoning, adversarial attacks, model evasion, and misuse of AI agents.
- Assist in modeling threats to AI agents and analyzing the security of Model Context Protocols (MCP), focusing on risks such as context leakage, corruption, unauthorized agent behavior, and external tool abuse.
- Monitor threat intelligence feeds, academic research, and emerging AI attack vectors to keep Cisco’s AI Defense teams informed of new risks.
- Contribute to writing high-quality internal intelligence reports, threat assessments, technical advisories, and research papers.
- Collaborate closely with engineering, security, product, and research teams to integrate findings into defense strategies and product hardening initiatives.
- Build small proof-of-concept demonstrations to validate attack theories against AI agents, context protocols, and model memory graphs.
Other
- Work closely with the Director and Tech Lead of AI Threat Intelligence.
- Collaborate with fellow researchers, data scientists, security engineers, and product teams.
- Interest in responsible AI practices and secure deployment of autonomous and agentic AI systems.
- WeAreCisco where every individual brings their unique skills and perspectives together to pursue our purpose of powering an inclusive future for all.
- Our passion is connection—we celebrate our employees’ diverse set of backgrounds and focus on unlocking potential.
