Iyiola Emmanuel Olatunji
Postdoctoral Researcher, University of Luxembourg
TruX Research Group — SnT
Secure and trustworthy AI for software engineering, privacy, and graph learning.
emmanuel [dot] olatunji [at] uni [dot] lu
I am a postdoctoral researcher in the TruX research group at SnT, University of Luxembourg, advised by Tegawendé Bissyandé and Jacques Klein.
Previously, I worked with Adam Dziedzic and Franziska Boenisch at CISPA. I received my Ph.D. in Computer Science from Leibniz University Hannover in 2024 under the supervision of Wolfgang Nejdl and Megha Khosla.
My research focuses on secure code generation, LLM-based software engineering and security, privacy-preserving machine learning, and graph learning, with an emphasis on building trustworthy AI systems that behave reliably under realistic threats.
News
- 2025 Joined TruX@SnT, University of Luxembourg as a Postdoctoral Researcher.
- 2024 Completed my Ph.D. in Computer Science at Leibniz University Hannover.
- 2024 Worked as a postdoctoral researcher at CISPA on privacy and security in modern AI systems.
- 2021 Received the Best Student Paper Award at IEEE Trust, Privacy and Security in Intelligent Systems and Applications.
Research
My research studies trustworthy AI for software, language, and graph-based systems, especially where failures create real security, privacy, or reliability risks.
Secure Code Generation and LLM Safety
Robust evaluation and defense for LLM-based software engineering.
I investigate how language models generate insecure or policy-violating code, how these failures can be evaluated realistically, and how to build stronger defenses, benchmarks, and assessment pipelines for programming assistants and code-generation systems.
Privacy, Security, and Graph Learning
Reliable machine learning under privacy and adversarial pressure.
I study privacy leakage, adversarial behavior, and robustness in machine learning systems, including graph neural networks, graph-aware LLMs, and privacy-preserving learning methods for sensitive data.
Prospective Students
I am happy to hear from students and collaborators interested in secure code generation, LLM security, privacy-preserving machine learning, graph learning, and trustworthy AI evaluation.
- Secure Code Generation Safer code synthesis, vulnerability-aware evaluation, and defensive prompting for LLM coding systems.
- LLM Security and Software Engineering Reliability, leakage, misuse resistance, and rigorous benchmarking for software-focused language models.
- Privacy-Preserving Machine Learning Learning with sensitive data, privacy risks, defenses, and dependable ML under strong constraints.
- Graph Learning and Trustworthy AI Graph neural networks, graph-aware LLMs, and robust learning behavior in high-risk settings.
If you would like to work with me, please email your CV, a short note on your research interests, and any relevant materials such as a transcript, paper, or project portfolio.
Publications
For a full and up-to-date list, see my Google Scholar.
- Secure code generation and LLM-based software engineering Selected work on secure and trustworthy code generation 2025
- Privacy and security in graph learning systems Research on privacy risks, defenses, and robustness in graph-based AI 2024
- Trustworthy AI evaluation under realistic threats Benchmarks and evaluation methods for modern AI systems 2024
- Privacy-preserving machine learning for sensitive data Models and defenses for privacy-aware learning systems 2023