Cami Rincón:

»The decision to articulate, measure and operationalise identity is neither neutral nor universally accepted as an appropriate practice of recognition. In turn, policy should not treat it as such.

Categories constituting technical identity such as race and gender are inherited from statistical standards rooted in colonial histories, which have been revised and adapted over time to maintain intersectional social hierarchies within modern constraints. Failing to intentionally revisit identity categories and categorisation practices within technical systems is bound to replicate, maintain and naturalise historical inequalities through system outcomes. This has already been extensively documented in a vast array of discriminatory outcomes from data systems that process identity.

As societies pursue increasingly sociotechnical trajectories, we must ensure that identity is managed in alignment with inclusive epistemologies. Inclusion must not further compromise those with marginalised identities through exposure within systems that do not attend to their needs and requirements. Rather, inclusion must be grounded in having agency over those systems. This entails supporting opting out of these systems, ensuring that digital identities can be shaped by those they represent so as to align with self- and community perceptions, and that digital identities benefit those they represent through their standards.

Designing for fluidity may enable this. Fluidity refers to agency-facilitating practices supporting the right to refuse, the right to opacity, the right to modify or delete data, and practices of self-identification. It entails no increase in either granular surveillance, nor further externally-prescribed categories. Individuals and communities must be supported in contesting the use of categorisations, articulating any categories that are used, scrutinising their benefits, and ultimately in declining to participate if their requirements are not met.«