Identification versus anonymity in e-government ANALYSIS

Analysis

Based on our analysis of current Belgian federal eGovernment, it seems obvious that it relies heavily on globally unique identification via globally unique identifiers and other authentic identity data that applies across several contexts.  In other words, Bentham’s words on identification and control (supra, section 2.4) are still very relevant in Belgium today: “who are you? With whom am I dealing? There [is] no room for evasion in answering this important question”.

What if changing the Benthamite identification scheme – apart from the contexts where the Belgian Privacy Commission agrees that this is indeed needed (namely in judicial and health matters) – is not a realistic option, for example because it would hinder the development of eGovernment? We explained earlier that privacy is not an absolute right, and there are other valuable interests, such as the public interest, which may limit the right to privacy, especially in a governmental context. 

The question we asked earlier is whether some degree of privacy could still be provided in case a government chooses for a Benthamite approach on identification and identity knowledge. We believe the answer to this question is yes. As will be explained in more detail in FIDIS deliverable D16.1, besides IDM systems that focus on obfuscation of the identity, for example via user-controlled context-dependent role and pseudonym management (FIDIS type 3 IDM systems), there are other legally admissible ways to incorporate privacy and data protection requirements in an IDM architecture used in eGovernment. One could, for example, take into account the model set out above, with prior authorizations to exchange data by a sectoral committee of the Privacy Commission.

From a privacy and data protection perspective, it is important that in the Belgian federal case, information is, at least in theory, only made accessible and exchanged with thereto authorized entities, based on formal authorizations by (a subcommittee of) the Belgian privacy commission. In sum, privacy enhancements and compliance with data protection regulation in such an alternative model is not gained via the empowerment of the user, but via control by a trusted third party.

The privacy enhancements we propose to make to this system are to make data processing of the personal data of their clients more transparent via monitoring and feedback procedures (transparency tools) and by incorporating basic data protection and privacy requirements in the basic architecture design. A potential “privacy-friendly” solution could then be to make the prior authorizations of the Belgian Privacy Commission enforceable via privacy policy enforcement. More precisely, data handling and data release policies could help enforcing authorizations of the above-mentioned trusted third party. This will be further elaborated in the deliverables of FIDIS workpackage WP16 on eGovernment.

In summary, through transparency, accountability, and enforcement mechanisms applied to the processing of “authentic” data about citizens in eGovernment, ‘data owners’ can regain more control, autonomy, self-determination, and thus privacy with regard to their data.