Fahad Shamshad
fahad.shamshad3 <at> gmail.com

I am a PhD student in the Computer Vision Department at MBZUAI, advised by Dr. Karthik Nandakumar and Dr. Salman Khan. My research focuses on safeguarding visual content in generative AI through privacy-preserving methods, robust concept erasure, and reliable watermarking. I develop defenses against malicious editing and unauthorized personalization to enable the safe and responsible deployment of generative models.

Email  /  Google Scholar  /  LinkedIn

I am open to research collaborations. If you are interested in working together on generative AI safety, privacy-preserving vision, or trustworthy AI, feel free to reach out via email or connect with me on LinkedIn.

News
  • Two papers accepted in CVPR 2026 (Main).
  • Selected for the SatML Travel Grant.
  • FaceGuardian accepted as Oral+Poster at SatML 2026.
  • TrojanWave accepted as Oral at EMNLP 2025.
  • Face Anonymixer accepted as a Spotlight Poster at IJCB 2025 (the flagship conference in biometrics).
  • Recognized as a CVPR 2025 Outstanding Reviewer Award (Top 5.6% of 13,000 reviewers).
  • Selected for the CVPR 2025 Travel Grant.
  • STEREO accepted as a Highlight at CVPR 2025.
  • Won the Best Poster Award at the MBZUAI Research Symposium 2024 (selected among 50+ PhD posters across ML, CV, and NLP).
  • Won the NeurIPS 2024 Invisible Watermark Removal Challenge across black-box and beige-box tracks among 140+ teams.
  • Our proposal to stress-test the DALL·E deepfake classifier was accepted by OpenAI under the DALL·E Detection Classifier Access Program.
  • Two papers accepted at MICCAI 2024: BaPLe and PromptSmooth.
  • Two papers accepted at CVPR 2023.
  • Book chapter published in Deep Learning for Medical Image Analysis, The MICCAI Society Book Series.