PhD Researcher: Privacy-Preserving Federated Learning

Turun yliopisto; Turku; On-site; EUR 29,000 - 36,000; Full time

Job summary

The University of Turku invites applications for a full-time Doctoral Researcher in the Department of Computing, Faculty of Technology, to work on Privacy-Preserving Federated Learning and Trustworthy AI. The position is fixed-term and ends 31.8.2029, with start by mutual agreement.

You will design self-learning federated frameworks, integrate PETs, and develop explainable AI components, contributing to secure model lifecycles and collaboration with academia and industry.

Qualifications

• MSc degree in Computer Science, AI, or Cybersecurity.
• Solid background in cybersecurity, ML, or LLMs.
• Familiarity with distributed computing, applied cryptography or XAI.
• Excellent analytical and problem-solving skills.
• Excellent English communication skills (IELTS/TOEFL).
• Programming skills in Python, C++, Rust.
• Interest in collaborative research with academia and industry.

Responsibilities

• Design and develop a self-learning Bi-Level FL framework that ensures global coherence while allowing local adaptability.
• Integrate Privacy Enhancing Technologies (PETs) such as Differential Privacy and secure aggregation.
• Implement selective unlearning mechanisms within the federated framework to eradicate outdated or biased knowledge, ensuring strict regulatory compliance.
• Develop a self-improving Bias & Explainability (XAI) agent that generates human-readable justifications for cybersecurity decisions and links them to system provenance.
• Contribute to model-level adaptive access control, utilizing cryptographic fingerprinting and Proof of Learning to guarantee the integrity of the AI model lifecycle.

Skills

Cybersecurity background

Machine Learning / LLMs

Distributed computing

Analytical thinking

Communication skills

Problem solving

Education

Master's degree in Computer Science / AI / Cybersecurity

Tools

Python

C++

Rust

Job description

The University of Turku invites applications for a full-time Doctoral Researcher in the Department of Computing, Faculty of Technology, to work on Privacy-Preserving Federated Learning and Trustworthy AI. The position is fixed-term and ends 31.8.2029, with start by mutual agreement.

You will design self-learning federated frameworks, integrate PETs, and develop explainable AI components, contributing to secure model lifecycles and collaboration with academia and industry.