Link Search Menu Expand Document

Project Publications

3.5.4 Anticipating legal protection under data- and code-driven normativity

By Tatiana Duarte

In text-driven normativity (TDN), legal reasoning and interpretation are purposeful practices whose object is constituted by an interplay of facts, practices and sources of law. The expected outcome of such interplay is a judicial decision. The human mediation that necessarily comes with interpretation engages human experience within the transfiguration of brute into institutional facts (more precisely, into legally relevant facts).1

This scenario is altered in data-driven technologies. Facts are reduced to data points processed according to a mathematical function which establishes relations between them with the objective of making predictions.

  1. To what extent does the compiling of human experience into data points fundamentally change legal reasoning and interpretation?
  2. How will the legal system maintain the self-generative nature ensured by TDN where human experience is datified?
  3. How will the system of countervailing powers look like in the moment of deciding the relevant data points to be the taken as the input of data-driven interpretation? In what terms can statistics be absorbed by the mode of veridiction of law?
  4. Can the vices (préjuges) of human reasoning and interpretation be amended by the virtues of a mathematical formula – or can it also be the other way around?
  5. Will the mathematisation (and the use of statistics) bring more certainty to legal reasoning and interpretation? Are there different kinds of certainty? If so, what are the trade-offs of different kinds of certainty within the system of countervailing powers?

CDN involves the creation of rules in programming language that self-execute when some conditions are met. Private companies develop and deploy digital architectures that are designed to serve the interests decided by the programmer or by the company who employs them.

The digislator (digital legislator) is not a democratic representative, nor they are addressable by the legal community as a legitimate author or reader of the social contract. The performative force of code in shaping human behaviour is a-legal2 as it happens out of the framework instituted by the rule of law.

  1. Is the current system of checks and balances translatable into code or in any way embeddable in the pipeline of automated enactment/enforcement? What will be the relation between enactment and publication of rules? How will the concepts of publication, validity and efficacy articulate in code-driven law?
  2. To what extent does the synchronic enactment and execution of legal norms make legal reasoning and interpretation obsolete practices?
  3. Can the principles of legal reasoning, such as the adversarial nature of procedure, be embedded in code-driven law? If so, in what terms?
  4. Do automated regulation and automated compliance demand different articulations of the system of countervailing powers? How would that articulation look like?
  5. Both legal interpretation and code require expertise to be understandable – and, thus, contestable. In what terms could CDN engage with contestability?
  6. If the source code is not published before deployment, individuals will only realise that rules are being applied in the moment of execution. Does CDN claim for new arrangements on the system of countervailing powers or none of such adjustments would ever make the application of code (as law) legitimate?

References

  1. C. Jabloner, ‘How the Facts Enter Into the Law’, in N. Bersier Ladavac, C. Bezemek and F. Schauer (eds.), The Normative Force of the Factual — Legal Philosophy Between Is and Ought, Law and Philosophy Library, Vol. 130 (Springer International Publishing 2019), pp. 97-110. ↩

  2. L. Diver, ‘Digisprudence: the design of legitimate code’ (2021) 13(2) Law, Innovation & Technology (forthcoming), p. 5. ↩

This page was last updated on 16 March 2023.