Privacy, Identity and Trust in Electronic Voting Technologies

点击翻译按钮获取中文摘要

This proposal addresses cybersecurity in the context of explosive interest and adoption of information and communication technology (ICT) into Canada's elections. Online voting, for example, is currently offered in almost half of Ontario's municipalities and expected to double by 2022. In October 2019, the Northwest Territories became the first sub-national election agency to offer online voting, and Quebec is currently holding a public consultation on this technology. Online voting is seen as a means of boosting voter turnout and engagement and building self-governance capacity in Indigenous nations. Despite this rapid growth and interest, however, no standards for online voting exist in Canada, and federal and provincial governments have warned of cybersecurity concerns. With the goal of providing Canada with a responsible path toward election ICT, the PI proposes to investigate cyber technologies that will protect the democratic process as we enter this new era of digital democracy. Elections as a cybersecurity design problem are highly constrained. Solutions must be inclusive to a highly diverse electorate, protect against an unusually broad threat landscape, and juggle the competing properties of ballot secrecy and public accountability. Developing ICT along these lines presents a unique research challenge requiring fundamentally novel techniques. The research objectives of this program will coincide with three pressing cybersecurity gaps in contemporary election ICT: trust, privacy, and identity. The trust objective of the proposed research program will focus on the technical challenge of providing an electorate with compelling, objectively verifiable evidence that an election result is correctly counted and reported. This will be accomplished through the design of novel techniques, algorithms, and protocols for cryptographic end-to-end verifiable internet voting (E2E-VIV), and the use of blockchains for specialized tasks such as automatic smart-contract based proof validation. The identity objective will tackle the broader issue of client-side security and voter eligibility to identify and demonstrate vulnerabilities in existing deployments to malware, phishing, cross-site scripting, and knowledge-based authentication. The privacy objective will focus on strengthening formal guarantees to ballot secrecy in connection with advances made under the confidence and identity objectives. This includes studying the privacy impact of existing confidence and identity ICT through the inter-disciplinary application of techniques such as re-identification risk from the health data privacy world. It will also explore the long-term privacy prospects of E2E-VIV proof transcripts through the development of post-quantum cryptographic techniques and protocols.