The Seat at the Table Where the Algorithm Mediates: How Technology, AI and Mediation are Interacting in the Indian Setting

With the number of cases pending before the Indian courts rising exponentially, artificial intelligence is making attempts to break into the realm of dispute resolution – but is Indian legislation ready to embrace this new trend? More than 50 million cases¹ are pending before
Indian courts, 2023 Mediation Act enacted in year and an 80 % decrease in time required in ODR process.

Introduction - AI Mediator? Not in the Distant Future

Imagine yourself logging into a platform, explaining your disagreement with a contractor in a few sentences and receiving, in mere hours, a well-structured offer to resolve the dispute, made not by a human mediator but an artificial intelligence algorithm. There are no meetings to attend, no long-distance trips to undertake, and no need to spend years awaiting trial – a process slowed down immensely by a pending backlog of cases numbering up to 50 million.

While this might sound far-fetched, the reality is quite different – platforms such as Presolv360, SAMA, and the project of the Supreme Court of India based on NALSA-ODR Initiative have already started implementing technology in India's dispute-resolution process. The question here is not whether AI algorithms are capable of entering the realm of mediation – they have already done so – it is only on what terms and under what conditions.

Legislation Background - Legislative Framework: From Arbitration to Mediation Act, 2023

India’s engagement with alternative dispute resolution predates the post-liberalisation legislative reforms by several decades. The Industrial Disputes Act, 1947² institutionalised conciliation and arbitration in labour disputes as early as 1947. In the domain of commercial disputes, the Arbitration Act, 1940³ governed domestic arbitration for over five decades before being repealed. The Legal Services Authorities Act, 1987⁴ then established the institutional framework for Lok Adalats; though enacted in 1987, the National Legal Services Authority constituted under it became operational only in 1995. The Arbitration and Conciliation Act, 1996⁵ — modelled on
the UNCITRAL Model Law on International Commercial Arbitration — then comprehensively overhauled the arbitration and conciliation framework, providing for the Principle of Party Autonomy and non-adjudicative dispute resolution. Subsequently, Section 89 of the Code of Civil Procedure, 1908⁶ introduced court-referred mediation as a method of dispute resolution; though its implementation across subordinate courts has been uneven, the Supreme Court of India has elaborated its scope through a body of jurisprudence beginning with Salem Advocate
Bar Association v Union of India (2005).

A landmark breakthrough happened when the Parliament adopted the first legislative act dedicated solely to mediation: the Mediation Act, 2023⁷. The law includes provisions concerning online mediation, mediation prior to the initiation of a civil suit, and enforcement of settlement agreements similar to court decrees. Significantly, neither does Section 3(h) of the Mediation Act exclude, nor does it regulate the use of artificial intelligence technologies in ADR in any way.

On the constitutional level, the right to seek justice under Article 21⁸, and the directive for equality of justice and availability of free legal aid under Article 39-A⁹, constitute the framework that allows for the introduction of technology-based ADR solutions, including those intended for marginalized section of society unable to engage in lengthy and costly litigation.

The Mediation Act, 2023, is a highly progressive step, but the lack of regulation regarding artificial intelligence is a flaw that needs to be addressed in the future by the legislature.

Debates and Controversies - Central Questions Surrounding AI-Mediated Solutions: Are Artificial Intelligence Algorithms Neutral Enough?

The spirit of mediation requires a mediator – an unbiased individual capable of reading the atmosphere in the room and guiding the process of resolving the dispute. Thus, questions concerning the ability of the algorithm to act as a neutral party arise while using AI tools in mediation. Firstly, can an algorithm act neutrally, considering its training based on historical data. In a country such as India, where conflicts often reflect the disparities in power, an AI trained to find solutions for the same kind of disputes in the past may only reinforce the imbalance of interests.

Additionally, can an individual give an individual give her/his their informed consent for entering into an ADR process to entering an ADR process while dealing with an algorithm instead of human mediator? And finally, how many privacy concerns privacy concerns arise when personal or corporate information processed by the algorithm in order to resolve the dispute?

Court Decisions - How the Judiciary Has Spoken (and Has Not Spoken) on the Issue So Far

Salem Advocate Bar Association v. Union of India (2005) – Supreme Court of India
Supreme Court supported Section 89 of the Code of Civil Procedure¹⁰, and emphasized that the means of Alternative Dispute Resolution are to be used genuinely and in earnest, rather than for mere formalities¹¹. By recognizing the role of mediation in delivering justice in India, the Court laid the basis for allowing the use of advanced technological tools in mediation in future cases.

Afcons Infrastructure Ltd. v. Cherian Varkey Construction (2010) – Supreme Court of India
In this Judgement, the Court defined the categories of disputes suitable for being resolved through ADR mechanisms and drew a line between the disputes where the parties may settle issues themselves and those related to the public interest or the violation of constitutional rights¹². This clarification has immense value when speaking about AI-mediated disputes.

Justice K.S. Puttaswamy v. Union of India (2017) – Supreme Court of India
In this landmark judgment delivered by a nine-judge bench, the right to privacy was recognized as one of the fundamental rights enshrined in Article 21 of the Constitution¹³. Hence, any platform processing data of the disputing parties during mediation must comply with constitutional regulations – something not currently ensured in any manner.

As of today, there are no High Court or Supreme Court decisions concerning the legality and admissibility of evidence generated through AI-mediated procedures.

Critical Analysis - Gaps and Problems That Cannot Be Ignored

The Mediation Act, 2023 prescribes that a mediator must be registered according to a recognized scheme and pass appropriate training.¹⁴ While it is evident that an AI platform may not be registered as a mediator, the question of its legal does arise. Does it mean that the mediator uses an AI platform as a tool, or that the platform acts as the mediator itself? The issue here is extremely important because any adverse effects caused by the use of biased AI-recommendations will require identifying and holding either the developer, the user or the mediator liable.

Additionally, there is an aspect of digital divide to be considered. Online Dispute Resolution requires sufficient digital literacy of all involved parties and a stable internet connection. For the semi-urban and rural India with numerous disputing individuals, this problem may prove to be a crucial challenge for the success of ODR programs.

More subtle than any of those issues might be the question of what gets lost when we take the human out of the room. The very process of achieving agreement through mediation – In the Indian context, the Lok Adalat system is itself premised on a similar philosophy of sulh (amicable settlement) and panchayat-style consensus, in which the relational dimension is central – could never be replicated by a machine. Whether an artificial intelligence could truly serve as a substitute for that relational component of mediation is not for the law alone to answer.

Contemporary Relevance - Policy Trends and the Highway Already Being Laid

The National Policy Plan for ADR¹⁵, the push towards ODR in consumer¹⁶ and commercial disputes by the NITI Aayog¹⁷, and the Supreme Court’s own e-court project indicate a clear governmental appetite for technological approaches to dispute resolution. Even the current Consumer Protection (E-Commerce) Rules encourage grievance redress using ODR mechanisms.

By 2024, NALSA and some High Courts had launched pilot projects using AI-based dispute categorisation systems to divert matters into relevant ADR forums.¹⁸ Though such systems support human discretion rather than supplanting it, they are nevertheless just one piece of the wedge that will continue widening.

Internationally, the EU’s AI Act (2024), labels dispute resolution AI as “high risk,” mandating transparency, human supervision, and auditing of biases.¹⁹ India lacks such a policy currently, though the ongoing consultations surrounding the Digital India Act may provide some solutions.

The urgency of the matter is not, of course, to stop technology in its tracks. It is to ensure the appropriate safeguards are put in place before the vehicle is already moving too fast to be steered.

Conclusion - An Open Verdict

The use of technology in mediation cannot be considered as inherently harmful or helpful, but rather a potentially powerful tool depending on its surrounding governance architecture. Indeed, India stands uniquely poised to establish such a framework preemptively, a rare occurrence in the annals of law-making. The Mediation Act, 2023 is a solid base. All it needs now is another level: concrete regulation around AI transparency, algorithmic responsibility, privacy in ADR proceedings, and, most importantly, assurances that efficiency will never take precedence over justice.

What may be the most critical question may not be whether AI is capable of conducting mediation. It may be whether we are wise enough to determine under what circumstances we should employ AI for such purpose, and brave enough to reject it when necessary.

Do you believe the Indian legislature should mandate its own framework for regulating AI in ADR procedures, or simply defer the issue to the courts as disputes arise?

References

1. National Judicial Data Grid, "Pendency Statistics"(Supreme Court of India, Department of Justice, 2023) <njdg.ecourts.gov.in> ↩
2. Industrial Disputes Act 1947 (India), ss 4–12 (conciliation officers, boards of conciliation, and arbitration). The Act predates the Legal Services Authorities Act 1987 and represents India’s earliest statutory institutionalisation of ADR in the labour context. ↩
3. Arbitration Act 1940 (India) (repealed by the Arbitration and Conciliation Act 1996). ↩
4. Legal Services Authorities Act 1987 (India). ↩
5. Arbitration and Conciliation Act 1996 (India). ↩
6. Code of Civil Procedure 1908 (India), s 89. ↩
7. Mediation Act 2023 (India). ↩
8. Constitution of India 1950, art 21. ↩
9. Constitution of India 1950, art 39-A. ↩
10. Code of Civil Procedure 1908 (India), s 89. ↩
11. Salem Advocate Bar Association v Union of India (2005) 6 SCC 344 (Supreme Court of India). ↩
12. Afcons Infrastructure Ltd v Cherian Varkey Construction Co (P) Ltd (2010) 8 SCC 24 (Supreme Court of India). ↩
13. Justice KS Puttaswamy (Retd) v Union of India (2017) 10 SCC 1 (Supreme Court of India) (nine-judge bench). ↩
14. Mediation Act 2023 (India), ss 13–14 (mediator registration and qualification requirements). ↩
15. NITI Aayog, Designing the Future of Dispute Resolution: The ODR Policy Plan for India (2021). ↩
16. Consumer Protection (E-Commerce) Rules 2020 (India). ↩
17. Supreme Court of India, Phase III e-Courts Project: Vision Document (Department of Justice, Ministry of Law and Justice 2023). ↩
18. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) [2024] OJ L 1689/1. ↩
19. Ministry of Electronics and Information Technology, Government of India, Digital India Act 2023: Consultation Paper (2023). ↩