Privacy-by-Design LLM Research Internship

Inria Palaiseau; Palaiseau; On-site; EUR 12,000 - 17,000; Full time

Job summary

Inria Palaiseau is offering an internship focused on privacy-preserving techniques for LLM-based systems handling user requests. You will study information flows, formalize leakage metrics, and explore mechanisms to reduce leakage while preserving output quality.

Aimed at candidates with strong Python skills and an interest in privacy, security, and networking. The project encompasses literature review, threat modeling, method design, and experimental evaluation against baselines, within a

Qualifications

• Interest in privacy and security; differential privacy a plus.
• Experience with LLMs and solid Python skills.
• Knowledge of networking is a plus.

Responsibilities

• Review privacy-related literature and threat models in LLM systems.
• Define privacy leakage metrics and candidate network use cases.
• Design and prototype approaches to reduce leakage while preserving utility.
• Evaluate leakage reduction techniques against baselines in experiments.

Skills

Python

LLMs

Privacy

Differential privacy

Networking basics

Job description

Inria Palaiseau is offering an internship focused on privacy-preserving techniques for LLM-based systems handling user requests. You will study information flows, formalize leakage metrics, and explore mechanisms to reduce leakage while preserving output quality.

Aimed at candidates with strong Python skills and an interest in privacy, security, and networking. The project encompasses literature review, threat modeling, method design, and experimental evaluation against baselines, within a