Nowadays the principle of fairness has become an increasingly important and widespread legal principle for the regulation of digital, algorithmic/data-driven technologies. This is particularly so at EU level where, among other examples, the principle can be found in the General Data Protection Regulation (GDPR), the Unfair Commercial Practices Directive (UCPD) -relevant for the regulation of dark patterns for instance, or even underpinning concepts of algorithmic fairness and bias (or absence thereof) that are also part of the upcoming AI Act.

Up until recently, the exact meaning of this principle has been elusive. However, there is increasing agreement to see it as partaking of the proportionality principle, and its technique of adjudicating conflicts between competing claims. This is certainly the case in data protection law (see EDPB, 2019), and also in consumer protection law (see, UCPD, Recital 6).

Taking the adequation between proportionality and fairness as a starting point, this contribution asks what can be expected from this all-encompassing use of proportionality/fairness. It answers it by (re)tracing a different genealogy of the principle. Contrary to usual accounts (that oftentimes go back to Aristotle and/or 19th century Germany), it argues that a different genealogy of proportionality can be traced. One that shares many commonalities with the concept of cybernetics, which is at the heart of information technology, which itself underpins the functioning of algorithms, and thus data-driven technologies.

By (re)tracing the cybernetical origin of the principle of proportionality/fairness, this contribution asks what it means to re-think the principle of fairness -and in turn the whole EU digital technology regulation apparatus- as a cybernetic practice, and what in turn this means for the type of protection that can be expected from it.