- Robust, explainable, and security-aware AI for networked and communications-enabled systems, including cyber-physical systems (CPS).
- Research thrusts:
- CPS security and anomaly detection in water/wastewater systems
- AI assurance for networked AI and agentic workflows
- Generative AI security: synthetic data, detection, and attribution
- Focus on adversary-aware anomaly detection, robust learning under distribution shift, and operational trust/interpretability.
- Develops model-agnostic assurance methods, secure learning pipelines, and guardrails for agentic AI in cyber operations.
- Advances detection/attribution for synthetic/adversarial content and defenses against concealment/fabrication attacks.
- Funding: NSF SaTC, NSF CPS, ONR, DARPA. Group building at ODU with strong mentoring and open-source outputs.
