THE DIGITAL MEDIATOR: POLICY FRAMEWORKS FOR INTEGRATING ARTIFICIAL INTELLIGENCE INTO ALTERNATIVE DISPUTE RESOLUTION

Abstract

The international landscape of Alternative Dispute Resolution (ADR) finds itself on the cusp of a great technological paradigm shift.  A successful mediation is always anchored entirely on the human element, such as a mediator’s ability to utilize emotional intelligence, empathy, and neutrality. However, the emergence of Generative Artificial Intelligence (AI) and its predictive power has firmly established itself in the mediation room. Algorithms are no longer passive “fourth party” instead, they have evolved into an active “fifth party agent” capable of guiding negotiations and suggesting settlement options independently. This paper examines the profound policy and regulatory gap created by AI advancement by focusing on procedural disclosure gap and algorithmic-ingestion paradox. While analyzing how the Large Language Models, (LLM) clashes with the key mediation principles under the EU AI Act, The UNCITRAL Model Law and domestic frameworks like Punjab ADR Act 2019 are also put to the test. What happens when a settlement under the Singapore Convention is mediated by an algorithm? This paper argues that the answer should worry us. Consequently, this paper argues that the integrity of the process cannot be compromised just for the sake of expeditious computational output. It proposes the solution by step-by-step framework positioned on algorithmic sandboxing, transparent disclosure rules, a mandatory human-in-the-loop standard to uphold the integrity of digital justice.