AI Researcher at Cisco in Milpitas, CA

Job description

The application window is expected to close on: 09/27/2026 Meet the Team The Cisco AI Software & Platform Group drives the development of groundbreaking generative AI applications. We build enterprise AI platforms and developer tooling across the company, integrating our networking, security, and observability portfolio with modern AI to power autonomous systems. Fueled by a passion for AI/ML, we strive to create a secure future for our customers. Our collaborative, agile team thrives on tackling complex challenges and delivering innovative solutions globally. The AI Threat Intelligence and Security Research team helps make AI safer to build, deploy, and use. We study emerging threats across AI models, applications, agents, and supply chains, then translate our research into practical protections—including red teaming, security tools, threat detections, provenance capabilities, runtime guardrails, and product innovations. We also advance the field through open-source projects, industry frameworks, technical publications, customer and partner engagement, and education. Your Impact In this role, you will analyze emerging AI threat vectors, build automated detections, and collaborate closely with other AI/ML teams, product managers, and core engineering teams to protect cloud-scale AI systems. You will: Conduct AI threat intelligence research to identify, track, and profile malicious actors and adversarial techniques targeting AI models, applications, and agentic workflows. Develop, evaluate, and maintain detection content —including heuristics, behavioral detections, and telemetry-driven analytics—for integration into customer-facing cloud products. ​ Research and assess security risks across the AI lifecycle, from model development and training through deployment and runtime use, including prompt injection chains, provenance, data poisoning, backdoors, and malicious dependencies. Conduct adversarial testing and red teaming across models, applications, and agentic deployments. Translate research into impact through evaluations, detection pipelines, security frameworks, open-source tools, and customer guidance. Analyze and synthesize large-scale threat telemetry using modern data platforms such as Snowflake and AI-assisted analysis tools. Partner with engineering and product leadership to translate adversarial research into automated, production-ready defense mechanisms inside Cisco Cloud Control. Minimum Qualifications Bachelor's degree plus 5 years of related experience , Master's degree plus 3 years of related experience , or PhD plus 1 year of related experience . Prior experience in one or more areas such as cyber threat intelligence (CTI), security research, vulnerability analysis, incident response, or detection engineering. Ability to derive actionable insights from threat intelligence, large datasets, and/or ambiguous signals to inform clear understanding of threat landscape Applied experience developing and maintaining detection content, including YARA rules, behavioral detections, or threat queries. Experience conducting OSINT-based threat investigations and producing clear, actionable intelligence reports; familiarity with dark-web research methods is a plus. Previous experience with threat intelligence platforms and/or evaluating adversarial risks in AI/ML systems, such as prompt injection, data poisoning, model backdoors, evasion, or agent and tool abuse. Preferred Qualifications Applied experience with agentic AI architectures, tool-calling protocols such as the Model Context Protocol, or AI security and guardrail frameworks. Proficiency in Python and SQL for research, data analysis, and automation. Experience using AI-assisted coding tools such as Claude Code, Cursor, or Codex to automate workflows and accelerate research, and improve investigations through scripts, workflows, or other tooling. Background in adversarial red teaming, blue teaming, security operations, or mapping threats and controls to frameworks s