DOE Developing AI for Privacy-Sensitive Datasets

The U.S Department of Energy recently announced $1 million in collaborative research funding to develop artificial intelligence and machine learning algorithms for privacy-sensitive datasets such as biomedical or personal healthcare information.

The project, “PALISADE-X: Privacy Preserving Analysis and Learning in Secure and Distributed Enclaves and Exascale Systems,” comes in response to congressional direction for DOE to expand its collaborative research efforts with the National Institutes of Health (NIH) in the areas of data and computation.  The two agencies share an interest in developing privacy-protecting AI and machine learning for “grand challenge datasets” that are the focus of the NIH Bridge2AI program and may hold the key to problems such as predicting the severity of COVID-19.

“Coupling privacy-preserving artificial intelligence and algorithms and DOE’s high-performance computers with NIH data will accelerate biomedical research,” said Barbara Helland, Associate Director for Advanced Scientific Computing Research, DOE Office of Science.

The PALISADE-X project was chosen through competitive peer review to assess the merit of the project, the suitability of its proposed budget, and the capability of its personnel. Funding was awarded to Ravi Madduri of Argonne National Laboratory and Kyle Halliday of Lawrence Livermore National Laboratory.

The project will be led by Argonne National Laboratory, working alongside Lawrence Livermore National Laboratory, the University of Chicago, the Broad Institute, and Massachusetts General Hospital.