Restricted Verifier: Using AI/ML to Limit Data Disclosure

Decentralized identity (DI) initiatives offered on a governmental level raise the possibility of forcing the system on participants, while not addressing those who do not wish to join the system. This raises the possibility of total exclusion of some people, not only from the services but from certain life processes. Moreover, this problem does not only concern those individuals who completely refuse to participate in a new digital identity initiative but also those who are not willing to share all of the required information for a service. The disclosure of personal information is a prerequisite in a number of cases (e.g., international travel) with or without a specific application [1]. However, a service provider or verifier still has a superior position over the data holder, as they can dictate what information they require to be shared with them in order to use their service. It is possible for a service provider to request more data than they actually need, and the user would be forced to either comply or be excluded. In the situation where similar services are provided by more than one provider, this is manageable, but in the case of governmental institutions or international travel, there is no choice. 

The thesis will contribute to the development of the restricted verifier component, addressing the challenge of excessive data requests by verifiers and enhancing user privacy in DI systems. As part of the thesis, an ML/AI model will be selected and utilized to restrict the verifier in a neutral fashion, by providing knowledge on whether the requested personal data by the verifier is truly necessary for a given use case. The work on this component will involve (i) identifying a suitable ML/AI model, (ii) building a dataset for the model training, (iii) integrating the module into a functional SSI system (codebase will be provided), and (iv) quantitative evaluation of the newly restricted verifier component. 

Sources to Consider: 

[1] Eman, N. "COVID-19 Smart Air Pass Card", Journal of Law and Humanities Sciences, vol.15.04, 2022, pp. 11-27. 

[2] A. Mühle, A. Grüner, T. Gayvoronskaya, and C. Meinel, ‘‘A survey on essential components of a self-sovereign identity,’’ Computer Science Review, vol. 30, pp. 80–86, 2018.

[3] . Čučko, Š. Bećirović, A. Kamišalić, S. Mrdović, and M. Turkanović, ‘‘Towards the classification of self-sovereign identity properties,’’ IEEE access, vol. 10, pp. 88 306–88 329, 2022.