Ramesh Karri Research Group

The group studies trustworthy hardware and develops security-aware CAD, verification, and metrics to detect and prevent hardware Trojans and other tampering. Researchers ran an AI Hardware Attack Challenge and analyzed how large language models can be used to insert exploitable hardware modifications, documenting attack mechanisms and mitigations. The team investigates AI-orchestrated malware and defenses, exemplified by research on autonomous LLM-orchestrated ransomware prototypes. They also co-lead platform efforts to accelerate privacy-preserving cryptographic hardware design and shared chiplet libraries to improve secure hardware prototyping.