SpotAI

The Secure, Private, Open and Trustworthy (SPOT) AI lab at MBZUAI, led by Nils Lukas. We work on secure, private, and strategic machine learning — including content watermarking, privacy-preserving inference, AI safety & security, and conversational agents with social intelligence.

We are hiring multiple postdocs, backed by two new external grants — (i) multi-agent systems with privacy for healthcare (with Praneeth Vepakomma and Preslav Nakov) and (ii) AI safety for smallholder advisory (with Salem Lahlou and Martin Takáč). We are looking for exceptional talent in multi-agent systems, self-improvement, AI safety and (differential) privacy (e.g. our ICLR'26 paper on DP-Fusion), with strong first-author publications at top venues (NeurIPS, ICML, ICLR, IEEE S&P, CCS, USENIX Security). You would have dedicated NVIDIA H200 GPUs, a collaborative, research-friendly environment and a highly competitive package. To apply, e-mail me (nils.lukas@mbzuai.ac.ae) a brief, technical overview of one of your favorite papers that you authored and why you find it interesting. Please prefix your subject line with [Multi-Agent Privacy] for the first project or [AI Safety Advisory] for the second.