Recap: ELSA Health Privacy Challenge 2026

The Health Privacy Challenge was back for CAMDA 2026. And thanks to Hakime Öztürk, ELSA can share the recap of the finale with you:

The 2nd edition Health Privacy Challenge has been completed! The number of CAMDA extended abstract submissions doubled in this edition, with contributions from both returning and new teams. Participants were invited to present their solutions during the Health Privacy Session at #CAMDA, as part of the ISMB 2026 Conference on July 16, 2025, in Washington, D.C.

This year’s CAMDA Health Privacy Challenge began with an invited keynote by Hyunghoon Cho on privacy-preserving genomics, highlighting key research questions and the broad range of applications and setting the stage for the challenge’s motivation.

Following the keynote, the Health Privacy Session opened up with an introduction by Hakime Öztürk, who introduced the challenge, a community-driven initiative focused on benchmarking privacy-preserving synthetic data for bulk and single-cell gene expression datasets. She highlighted a shift in blue-team submissions toward foundation model–based architectures. Despite this trend, the privacy–utility trade-off remained apparent, particularly in terms of preserving biological utility. This edition also attracted more red-team submissions in both the bulk and single-cell tracks.

Following, challenge participants showcased their work:

  1. A Unified Weighted-Distance Framework for No-Box Membership Inference Attacks on Synthetic Gene Expression Data. Ruixuan Liu, Emory University, U.S.A.
  2. Privacy Auditing of Synthetic Single-Cell RNA-seq Data. Steven Golob, University of Washington, Tacoma, U.S.A.
  3. Does Synthetic Bulk RNA-seq Data Protect Donors? Privacy Auditing through Membership Inference Attacks. Charlene Jarrell, University of Washington, Tacoma, U.S.A.
  4. Privacy-Preserving Synthetic Single-Cell RNA-seq via NMF-Compressed Truncated Vine Copulas. Andrew Wicks, DFKZ, Germany
  5. From Graphical Models to Foundation Models: Synthetic Bulk RNA-seq Data Generation. Elias Chaibub Neto, Sage Bionetworks, U.S.A.
  6. Benchmarking Recent Single-Cell Synthetic Data Generation Methods on the OneK1K Dataset. Buse Giledereli, Boğaziçi University, Turkey
Participants of the Health Privacy Challenge at CAMDA conference 2026
Presenting at the Health Privacy Challenge at CAMDA 2026: Ruixuan Liu, Buse Giledereli, Steven Golob, Andrew Wicks, Charlene Jarrell, Jonathan D H Kim, Emma Szebeny . Not on the photo: Elias Chaibub Neto.

Two participants from the challenge received CAMDA prizes, selected from presentations across the entire CAMDA track, where participants from multiple competitions presented their work. Congratulations to Charlene and Ruixuan on this recognition!

Congratulations!

As ELSA, we want to thank all participants for joining the challenge and the organizers.