Boston University researchers Mayank Varia, Vasiliki Kalavri, and John Liagouris have received three new grants supporting research aimed at making sensitive data and artificial intelligence more secure and privacy-preserving.
The largest, a three-year, $1.2 million award from the National Science Foundation’s Cybersecurity Innovation for Cyberinfrastructure (CICI) program, will support the development of the Open Privacy-preserving Analytics Lab (OPAL), an open-source platform for conducting secure scientific research across public and commercial cloud environments.
OPAL will use advanced privacy-enhancing technologies, including secure multiparty computation and trusted execution environments, to protect data while it is stored, transmitted, and actively used. Designed for researchers working with sensitive information in fields ranging from medicine and education to the social sciences, the platform will make it easier to test and deploy privacy-preserving applications at scale.
“The funding will enable us to build the first-ever secure testbed for fundamental research on sensitive data, opening the door to new scientific discoveries while fully respecting the privacy regulations that govern this space,” said Kalavri, an assistant professor of computer science. “We are truly excited about what lies ahead.”
Liagouris also received an Amazon Research Award in Cryptography to advance secure multiparty computation, or MPC, a technology that allows multiple parties to jointly analyze data without revealing their individual information. The project will address key technical barriers that have limited the technology’s adoption, translating recent advances in cryptography into scalable systems that can support new real-world applications.
A third grant, from the Foresight Institute, will support Liagouris and Varia’s development of a decentralized platform for creating privacy-preserving training datasets for specialized AI models. The project will combine secure multiparty computation with differential privacy, allowing organizations to contribute proprietary or sensitive information to shared datasets while maintaining control over how their data is selected, processed, and disclosed.
“These awards represent a tremendous opportunity for our group and a great example of what collaborative research can achieve,” said Liagouris, an assistant professor of computer science. “Over the past six years, we have built a strong foundation in secure data processing by leveraging advanced cryptographic techniques that unlock applications not possible today. The new funding will allow us to bring that work directly to the stakeholders who need the technology most.”
Together, the three projects build on years of collaborative research at BU’s Complex Analytics and Scalable Processing Systems Lab (CASPLab) and bring together expertise in cryptography, distributed systems, data analytics, and AI. The work could help researchers and organizations collaborate on sensitive data while preserving privacy and intellectual property—and expand the kinds of data that can safely be used for scientific discovery and AI.