EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

The Legal and Regulatory Challenges Posed by the Integration of Predictive Analytics (PA) into International Arbitration

Anisha Sharma

Law, Economics and Society · 2025

DOI: 10.30560/les.v1n2p68 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Law, Economics and Society, 2025. Verfasst von Anisha Sharma. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This research examines the rapid adoption of predictive analytics in international ADR. It argues that although the technology produces efficiency gains - reducing costs and timelines while enhancing consistency of decision making - the very efficiency gains carry with them collective and individual legal and ethical risks. This analysis supports the argument that the PA's dependence on complicated, proprietary ML models results in an inescapable legitimacy crisis. The emergence of this crisis can be traced back to fundamental process disputes, particularly the "Black Box" defect, which introduces an insurmountable opacity to algorithms and acute flaws in data governance surrounding confidentiality, reverse engineering risk, and non-compliance with regimes like the GDPR. All of these violations contravene the principles of due process, ultimately jeopardizing the global enforceability of arbitral awards under the New York Convention. The key contention raised in the proceedings is the tension between the confidentiality of technical information and the non-derogability of a party's right to contest crucial inputs affecting the decision. Due to the lack of regulations, proprietary systems can operate without oversight. These systems will import systemic bias from historical data into future legal outcomes. This structural deficit creates a compelling ethical reason for requiring immediate regulation. The current failure of institutional control over a third-party vendor's technology necessitates a shift from simple ethical advice to enforceable and transnational standards, ensuring the long-term legitimacy of the international ADR system. This research proposes a comprehensive three-pillar regulatory framework to shift the ecosystem from opaque behaviour to auditable transparency. The framework requires (1) the tribunal to disclose the model specification used and its methodology mandatorily; (2) conducting a data provenance, quality and bias audit of the Data Protection model, that is not conducted by the software vendor, of all commercial PA models; and (3) mandating algorithmic contestability, where any parties has the right to contest the outcome of an algorithmic output through a forensic analysis carried out by an expert. Future research needs to look at empirical analyses assessing the relationship between PA use and award challenge rates. Additionally, future research should develop open-source, auditable PA models that non-commercial institutions can utilize. Ultimately, effective governance of algorithmic arbitration relies on international institutions, such as UNCITRAL, to incorporate transparency and accountability into the arbitral procedure in a meaningful manner.

Zitieren

APA

Sharma, A. (2025). The Legal and Regulatory Challenges Posed by the Integration of Predictive Analytics (PA) into International Arbitration. Law, Economics and Society. https://doi.org/10.30560/les.v1n2p68

Vancouver

Sharma A. The Legal and Regulatory Challenges Posed by the Integration of Predictive Analytics (PA) into International Arbitration. Law, Economics and Society. 2025. doi:10.30560/les.v1n2p68

BibTeX
@article{Sharma2025,
  author = {Sharma, Anisha},
  title = {The Legal and Regulatory Challenges Posed by the Integration of Predictive Analytics (PA) into International Arbitration},
  journal = {Law, Economics and Society},
  year = {2025},
  doi = {10.30560/les.v1n2p68}
}
Kostenlos starten →