Research

Agentic AI security and resilient systems

The current program connects autonomous-agent security with long-running work on blockchain, cyber-physical systems, and software engineering.

Current direction

Autonomous agents increasingly act across model services, smart contracts, digital markets, and physical infrastructure. The research question is direct: how can these systems be audited, constrained, and kept resilient under adversarial conditions?

Agentic AI security and resilience

Current work examines how autonomous agents and multi-agent systems can be audited, secured, and made resilient when they interact with smart contracts, digital markets, and critical infrastructure.

  • Agent verification and security auditing
  • Adversarial and cooperative multi-agent learning
  • Decentralized AI and secure agent infrastructure
  • Security requirements for autonomous systems

Blockchain security and digital economies

Research on privacy, censorship resistance, maximal extractable value, decentralized finance, and protocol security in programmable blockchains.

  • Blockchain privacy and stealth addresses
  • Censorship resistance and protocol trust
  • DeFi security, MEV detection, and market behavior

Cyber-physical and infrastructure security

Security for smart grids, microgrids, connected vehicles, and other systems where software and AI act on physical infrastructure.

  • Cyber-resilient power-system protection
  • Attack detection for smart grids and industrial systems
  • Threat modeling for connected and IoT systems
  • Secure decentralized energy systems

Software and requirements engineering

Requirements and systems engineering provide the methods used to study trust, security, and failure in complex digital systems.

  • Security and privacy requirements
  • Requirements engineering for decentralized systems
  • Software architecture and architecture recovery