Privacy-preserving detection of online misinformation

This project will adapt and extend the functionality of PriPA, a state-of-the-art browser-based tool designed for Privacy-Preserving Analytics of language, which was developed between 2021 and 2022 on the AHRC/UKRI-funded Coronavirus Discourses project.

The new tool will use natural language processing techniques to enable the detection and flagging of linguistic misinformation markers to end users, whilst preserving end-to-end encryption to protect the privacy and security of their online browsing and communication activities. The tool will be available for researchers and individuals to download and use.

The tool will offer users privacy enhanced data processing for the detection of misinformation markers, and the opportunity to submit findings as participants in follow-on privacy preserving research. We will be working with consumers of online (mis)information to assess trust and efficacy of the tool.

By providing a way for automated systems to analyse data without undoing progress in end-to-end encryption, this tool will allow researchers to analyse the consumption of misinformation without infringing on the privacy of the participants.