01
Agent assurance and multi-agent security
Research on how autonomous agents can be audited, constrained, and kept resilient when they interact with other agents and external systems.
Research questionsforming · Abu Dhabi · Distributed
How can autonomous systems be trusted when they act across code, markets, and physical infrastructure?
Trust Lab studies the security and resilience of autonomous agents, multi-agent systems, programmable blockchains, and cyber-physical infrastructure.
- 01RequirementDefine the protected outcome
- 02Agent actionObserve plans and execution
- 03System effectMeasure expected and adverse change
- 04EvidencePreserve what can be challenged
Evidence revises the next requirement
Research
Questions before claims.
02
Decentralized systems and digital economies
Work on privacy, censorship resistance, maximal extractable value, protocol trust, and secure infrastructure for decentralized systems.
Research questions03
Cyber-physical resilience
Security for power systems, microgrids, connected devices, and other settings where software decisions affect physical infrastructure.
Research questions04
Requirements and systems engineering
Methods for specifying trust, recovering architecture, modelling threats, and examining failure in complex autonomous and digital systems.
Research questionsSelected evidence
Research records, not decorative claims.
MEV Detection for Ethereum Security: An Empirical Analysis on Agreement and Scalability
IEEE Transactions on Emerging Topics in Computing
Inspect publicationAn AI-Based Supervisory Measurement Integrity Validation Layer for Cyber-Resilient AC/DC Protection in Inverter-Based Microgrids
IEEE Transactions on Instrumentation and Measurement
Inspect publicationOn the Security and Privacy Implications of Large Language Models: In-Depth Threat Analysis
IEEE Cybermatics 2024
Inspect publicationMethod
Shared research standards
Specify trust before testing it
Each project states what a system must protect, who can challenge it, and which failures count.
Study systems under pressure
Security claims matter only after adversarial conditions, partial failure, and conflicting incentives enter the model.
Connect methods across fields
Agent assurance draws on security, requirements engineering, distributed systems, and empirical measurement.
Keep the evidence public
Papers, code, data, protocols, and negative results should remain inspectable whenever publication rules allow it.
Lead researcher
Davor Svetinovic
Lab lead · Professor of Computer Science and Associate Chair for Graduate Studies
Khalifa University · Visiting Fellow, ADIA Lab
Research leadership remains connected to public scholarly records and independently stated affiliations.
People and affiliations