Agentic AI Security
Studying threats to AI agents and developing defenses for secure tool use, planning, and autonomous decision-making.
Trustworthy & Secure AI LabSharif University of Technology
Trustworthy
and Secure AI.
We study the risks of machine learning—and explore how to build AI systems we can trust.
01 / The lab
About the Lab
The Trustworthy and Secure AI Lab is based in the Department of Computer Engineering at Sharif University of Technology. Our research focuses on the security, privacy, and trustworthiness of machine learning systems. We study how these systems can fail or be attacked, how sensitive information can be exposed, and how principled defenses and careful evaluation can improve their reliability in practice.
LocationRoom 503, Department of Computer Engineering
Sharif University of Technology, Tehran, Iran
02 / Research directions
Research Topics
Multi-Agent System Security
Investigating attacks and defenses in multi-agent systems, with a focus on secure communication, coordination, and collaboration.
Automated AI Red-Teaming
Developing automated methods to discover vulnerabilities, evaluate safeguards, and assess the security of AI systems.
03 / Our team
People
Principal Investigator
Portrait
Amir Mahdi Sadeghzadeh
Amir Mahdi Sadeghzadeh
Amir Mahdi Sadeghzadeh
Assistant Professor
Department of Computer Engineering
Sharif University of Technology
I lead the Trustworthy and Secure AI Lab, where we study the security, privacy, and trustworthiness of artificial intelligence systems.
PhD Students
Student portrait
Arman Shamsi
PhD Student
AI Agent Security
Student portrait
Zahra Arani
PhD Student
Jailbreak Multimodal Genrerative Models
Master's Students
Student portrait
Sajjad Vaezi
Master's Student
Multi-Agent systems Security
Student portrait
Javad Hezare
Master's Student
Understanding Refusal Mechanism in LLMs
Student portrait
Arian Komaei
Master's Student
Unlearning Methods
Student portrait
Arshin Kafi
Master's Student
Automated AI Agent Red-Teaming
Undergraduate Students
- [Student Name] — [Research topic or project]
04 / Research output
Publications
[Paper Title]
[Authors]
[Conference or Journal], [Year]
[Add links to the paper and code repository]
05 / Learning
Teaching
Security and Privacy in ML
2026 · Sharif University of Technology
06 / Work with us
Join the Lab
We welcome interest from students who are curious about the security, privacy, and trustworthiness of artificial intelligence. Explore our research topics and publications to learn more about the questions we study.
Prospective graduate students should consult the official admission requirements of Sharif University of Technology and the Department of Computer Engineering. Undergraduate students interested in research projects are also welcome to explore our work.
[Add the lab's contact email, application instructions, and current openings here.]
Frequently Asked Questions
How can I apply to join the lab?
Review the university's admission requirements and the lab's research interests. [Add the application process, whether prospective students should email the supervisor, and the materials to submit.]
How often do students meet with the supervisor?
[Specify the frequency and duration of individual meetings, and how students can ask for help between meetings.]
How are new research projects developed?
[Describe how students propose ideas, receive feedback, and discuss projects during group or individual meetings.]
How does collaboration work in the lab?
[Describe opportunities for collaboration within the lab and with external researchers, and how members support each other's work.]
What support does the supervisor provide during a project?
[Describe support with problem formulation, experiment design, interpretation of results, and manuscript preparation, together with the student's responsibilities.]
What are the expectations regarding progress and deadlines?
[Describe expectations for independent work, progress updates, planning, and communication about obstacles or delays.]