ABSTRACT
This article explores whether artificial intelligence (AI) in mediation enhances human dialogue or risks replacing the mediator. It analyses the growing use of AI in Alternative Dispute Resolution (ADR) through Online Dispute Resolution (ODR) platforms, where it improves efficiency, accessibility, and objectivity. In Kenya, where accredited mediators have increased from 74 in 2016 to 1,515 in 2025, full automation poses risks of de-skilling and loss of livelihood within a developing profession. While AI can process data and suggest solutions, it lacks the empathy and emotional intelligence essential to mediation. The paper argues for a hybrid, human-in-the-loop model in which AI supports rather than replaces mediators, ensuring fairness, accountability, and the preservation of mediation’s human essence through clear regulatory safeguards.
Keywords: AI, NLP, Mediation, ADR, ODR, Human-in-the-Loop, Kenya.
- INTRODUCTION
The age of artificial intelligence (AI) has arrived quietly but decisively, threading its way into corners of life that were once regarded as exclusively the domain of human judgment. From diagnosing illnesses to preparing legal briefs, it is now at the doorway of one of our most human practices; the art of dispute resolution. In its simplest form, AI refers to a system’s ability to correctly interpret external data, to learn from such data, and to use those learnings to achieve specific goals and tasks through flexible adaptation.[1] Many of these systems use natural language processing (NLP); the capacity to comprehend and generate human speech and data analysis to guide decisions.[2]
These tools are currently being tested in alternative dispute resolution (ADR), particularly mediation, within the legal sphere.[3] AI is being integrated into online dispute resolution (ODR) platforms to expedite negotiations.[4] Startups such as Bot Mediation, which just debuted at the 2025 ABA Techshow, claim to resolve conflicts in days rather than months by utilizing algorithms to produce neutral settlement solutions based on case facts.[5]
Kenya has not been left behind in this global transition. Despite recent controversy about AI-generated submissions with fictional authorities, the Kenyan judiciary is developing an AI Adoption Policy Framework to regulate the use of AI in legal practice.[6] Chief Justice Martha Koome has emphasised that the goal is to increase efficiency in case management, legal research, and administrative support while preserving judicial independence and data protection.[7] This initiative places Kenya as one of the region’s early adopters, mirroring broader arguments across the Global South about how AI can transform justice delivery.
In light of this, this article investigates whether AI in mediation should be viewed as a tool for improving dialogue or as an advancement that could potentially replace the mediator. It first evaluates the viability of AI in online dispute resolution, as well as its potential benefits in terms of efficiency, objectivity and accessibility. It then delves into the ethical and professional challenges that arise, before turning to Kenya’s growing regulatory approach and the human-in-the-loop model. The article concludes by proposing that AI should complement, not replace, the mediator in order to preserve the human core element of dispute resolution.
- CAN AI BE USED IN ODR AND WHAT’S ITS VIABILITY
AI can read vast amounts of negotiation and legal data in seconds, radically improving efficiency in mediation. For instance, Kenyan experts note that AI can rapidly browse through hundreds of legal documents for best solution-finding and due diligence. Globally, experts say generative AI can add and contrast large text sets in seconds; a feat that humans would take much longer to accomplish.[8] Through accelerating the review of documents and making negotiations easier, such software allows for disputes to proceed faster. Kenyan innovators echo this trend. Case in point, the Nairobi’s Utatuzi Center reports resolving cases in just 2–3 online sessions; cutting resolution time and costs by roughly 70%[9] while a U.S. startup, Bot Mediation, similarly claims to conclude mediations in days rather than months.[10]
Besides time efficiency, ODR dramatically widens access. It allows Utatuzi Center’s platform to provide automated dispute tracking 24/7 and secure video conferencing services eliminating the requirement of in-person meetings and making ADR usable across distances.[11] Utatuzi emphasizes the concept of the “3A’s”: affordable services, assisted legal service providers, and around-the-clock access to help vulnerable and remote communities participate in the process.[12] Likewise, the founders attest that their AI mediator is available at any time and, unlike traditional legal fees, negotiates them down to a fraction so as to enable mediation for those that could otherwise never afford it. Indeed, Utatuzi goes on to explain that bypassing bottlenecks in the courts translates into a huge time and cost saving for the parties, while maintaining confidentiality and thereby delivering justice on a broader scale.
AI also enhances objectivity in mediation. AI is not affected by emotions or personal biases, resulting in a more fair outcome.[13] In practice, AI can provide proposal recommendations founded on similar precedent cases, accelerating negotiations with evidence-based proposals. AI chatbots also aid in communication: people have requested features like ChatGPT to rewrite aggressive or accusatory language into more soothing, constructive language or even practice tough conversations in advance.[14] These amenities help with maintaining interactions respectful and terms of agreement fact-based, which would lead to parties being more secure in the process.
AI, has and continues to demonstrate its capacity to streamline aspects of dispute resolution such as automating scheduling, analyzing vast volumes of data and even suggesting neutral solutions to some disputes.[15] Yet, the question of whether it can replace the mediator is one that strikes at the very core of mediation itself.
- WILL AI REPLACE MEDIATORS?
Presently, the prospect of AI replacing the mediator is limited by both technical and ethical barriers.[16] While AI can process large volumes of information better than the human mediator it currently does not have the capability to understand human emotions.[17] This coupled with the risks of breaches, misuse of sensitive information,[18] which undermines the confidentiality of the session as well as the fabrication of evidence ideas or claims,[19] tilts the scales in favour of the human mediator leaving AI to best serve as a support tool rather than a substitute for the mediator.
However, technological innovation is inevitable and with it the question of AI replacing the mediator becomes more pressing. Current developments in AI aim to make the systems more human-like through advancements in engagement, collaboration, emotional understanding and social interaction.[20] These advancements seek to enable AI to sustain natural conversations, facilitate cooperation between conflicting parties, register and respond emotional cues,[21] and build rapport with users. Such progress suggests a future where AI could detect rising tensions between parties and intervene accordingly.
In Kenya the mediation sphere is steadily growing, with the number of accredited mediators rising from just 74 in 2016,[22] to 1,515 in 2025,[23] most of whom are actively engaged in court annexed mediation. If AI were to replace the mediator, then all these people face the risk of de-skilling and the potential loss of livelihood, undermining a system that has just begun to flourish. Such a situation calls for a clear regulatory framework governing the use of AI in alternative dispute resolution to ensure that the delicate balance between the human and technological aspects of dispute resolution is maintained.
- FUTURE OF AI IN ODR
This then begs the question, what role will AI play in the future of Online dispute resolution? Thus far, one thing that is established is that in a field characterised by person to person conversation, AI will exert its impact positively in the mediation process.[24] While this is the case, we can assess it as a double edge sword. In assessing conflict resolution, in the future, we are likely to see a hybrid approach. This is where AI will act as a virtual mediator using Natural Language Processing(NLP) and sentiment analysis. It will be able to interpret the emotions and intention of the parties involved, identifying points of contention and guiding the parties through a structured resolution process.[25]This reduces the burden on human mediators allowing them to focus on cases that require more nuanced judgement, harnessing both strengths to elevate and improve the mediation process to deliver better outcomes.[26] Developing it as a complimentary tool will lead to increased efficiency ie reducing the duration the amount of time mediation takes to fewer hours.[27] Additionally, features such as video conferencing can be utilized for real time discussion and utilization of AI algorithms so as to reach a resolution.[28]
On the other side of the sword lies one key aspect, the human-in-the loop principle. The Human-in-the-loop approach incorporates human decision oversight in AI decision making pipelines in order to uphold ethics and control.[29] The strategies for its implementation include: enabling a transparent design whereby humans monitor, explain and document AI behaviours for accountability purposes.[30] Second, Judicious Hybrid modelling allows for the humans to apply contextual intelligence where AI supplies data driven pattern recognition.[31] Lastly, humans provide continuous feedback to ensure rapid refinements against AI’s unusual patterns.[32] Therefore by ensuring we consider both sides of the sword, we are able to create a justice system that integrates technology’s strength while still considering integrity and humanity.[33] By harnessing it, we get better outcomes in the mediation process and more equitable outcomes, making justice more accessible.[34] Focusing on the human in the loop principle prevents AI-solutionism which is the tendency to believe that AI tools, by the mere fact that they exist, can resolve human centred disputes.[35] In response, AI must simply be used to support human judgement and not replace it.[36]
In Kenya, the judiciary is taking a step in the right direction towards the inclusion of AI in the integration of technology to automate court processes ensuring greater efficiency, transparency and access to justice.[37] This has been done through the first ever Judiciary Artificial Intelligence Adoption policy framework which just completed its on-going collection of comments and feedback in May 2025. The policy framework included AI integration in AJS systems such as mediation.[38] This integration allows for Online dispute resolution platforms that resolve disputes virtually and automated scheduling that organises mediation sessions efficiently.[39] Its relevance to the Social transformation through access to justice framework is that it closes the gap by expanding AJS and enhancing court annexed mediation by making them faster and inclusive by reaching marginalised communities.[40] Additionally, it includes the human in the loop principle whereby judicial officers and judges will maintain oversight over any AI driven recommendations. This will ensure judicial authority that is enshrined under Article 159 of the Constitution is upheld.[41]
- CONCLUSION
AI is transforming online dispute resolution by making negotiations more efficient. Bot mediation has demonstrated how technology can offer solutions within hours. However, developments in AI technology raise ethical and technical concerns, including the possible replacement of the mediator in the dispute resolution process. As Kenya’s mediation framework grows, AI must complement and not replace the mediator.
Article by SDRC :
- Asma Mohamed
- Charlotte Mwathi
- Ivan Kiriga
- Sagal Ahmed
[1] Christoph Bartneck, Christoph Lütge, Alan Wagner & Sean Welsh, An Introduction to Ethics in Robotics and AI (Springer, Cham, 2021) https://link.springer.com/book/10.1007/978-3-030-51110-4 on 29 September 2025.
[2] Diksha Khurana, Aditya Koli, Kiran Khatter & Sukhdev Singh, “Natural Language Processing: State of the Art, Current Trends and Challenges,” Multimedia Tools and Applications (2022) https://link.springer.com/article/10.1007/s11042-022-13428-4 on 1 October 2025.
[3] Katie Shonk, “AI Mediation: Using AI to Help Mediate Disputes,” PON – Program on Negotiation at Harvard Law School, 10 June 2025, https://www.pon.harvard.edu/daily/mediation/ai-mediation-using-ai-to-help-mediate-disputes/ on 1 October 2025.
[4] Francesca Tabor, ‘The AI Mediator: Can AI Help with Dispute Resolution and Peace-Keeping’ Francesca Tabor, 3 June 2025 – https://www.francescatabor.com/articles/2025/6/3/the-ai-mediator-can-ai-help-with-dispute-resolution-and-peace-keeping on 1 October 2025.
[5] Tabor, ‘The AI Mediator’.
[6] The Online Kenyan, “Kenyan Judiciary Adopts AI for Transcriptions as Google Enhances Gemini,” The Online Kenyan, 14 August 2025, https://www.theonlinekenyan.com/daily/2025-08-14/kenyan-judiciary-adopts-ai-for-transcriptions-as-google-enhances-gemini on 1 October 2025.
[7] The Online Kenyan, “Kenyan Judiciary Adopts AI for Transcriptions.”
[8] Using AI to Help Mediate Disputes.” (2025, June 10). PON (Program on Negotiation), Harvard. https://www.pon.harvard.edu/daily/mediation/ai-mediation-using-ai-to-help-mediate-disputes
[9] Getting Disputes Resolved Online by Utatuzi Center.” (2021, December 2). Communitas Africa. https://communitas.africa/2021/12/02/getting-disputes-resolved-online-by-utatuzi-center
[10] Using AI to Help Mediate Disputes.” (2025, June 10). PON (Program on Negotiation), Harvard. https://www.pon.harvard.edu/daily/mediation/ai-mediation-using-ai-to-help-mediate-disputes
[11] New Kenyan ODR Provider: Utatuzi Center.” (2021, February 10). National Center for Technology & Dispute Resolution. https://odr.info/new-kenyan-odr-provider-utatuzi-center
[12] Getting Disputes Resolved Online by Utatuzi Center.” (2021, December 2). Communitas Africa. https://communitas.africa/2021/12/02/getting-disputes-resolved-online-by-utatuzi-center
[13] Using AI to Help Mediate Disputes.” (2025, June 10). PON (Program on Negotiation), Harvard. https://www.pon.harvard.edu/daily/mediation/ai-mediation-using-ai-to-help-mediate-disputes
[14] Revolutionary AI-Powered Legal Dispute Resolution Solution Bot Mediation Selected to Present at ABA Tech 2025’s Pitch Competition.” (2025, March 28). LawNext. https://www.lawnext.com/2025/03/revolutionary-ai-powered-legal-dispute-resolution-solution-bot-mediationtm-selected-to-present-at-aba-tech-2025s-pitch-competition.html
[15] Levin B, ‘Will AI Replace Mediators? Debunking the Myths and Highlighting Collaborations’ MediatorLocal, 18 August 2025 — https://mediatorlocal.com/will-ai-replace-mediators/.
[16] Levin B, ‘Will AI Replace Mediators? Debunking the Myths and Highlighting Collaborations’ MediatorLocal, 18 August 2025 — https://mediatorlocal.com/will-ai-replace-mediators/.
[17] Alsamhan E, ‘AI and online dispute resolution: Mediation’ 4(13) Journal of Scientific Development for Studies and Research, March 2023, 283 https://orcid.org/0000-0001-6227-2667
[18] Garg M, ‘AI and mediation: A threat or helpful tool for mediators – an Indian perspective’ Jus Corpus Law Journal, 2023, https://www.juscorpus.com/ai-and-mediation-a-threat-or-helpful-tool-for-mediators-an-indian-perspective/ .
[19] Fasanya C, ‘Using artificial intelligence (AI) in ADR: Opportunities and challenges’ University of America, Curaçao, 2020
[20] Artificial intelligence: Robots, avatars, and the demise of the human mediator’ 25 Ohio State Journal on Dispute Resolution, 2010, 11.
[21] Zongni Li,Kin-Yeung Wong,Chan-Tong Lam,Attention-Based Transfer Enhancement Network for Cross-Corpus EEG Emotion Recognition, Sensors, 25, 18, (5718), (2025).
[22] ‘Judiciary establishes mediation suites to ease backlog’ Kenya News, 18 October 2023 — https://www.kenyanews.go.ke/judiciary-establishes-mediation-suites-to-ease-backlog/.
[23] ‘Court-Annexed Mediation’ Judiciary of Kenya — https://judiciary.go.ke/court-annexed-mediation/.
[24] Wäehlisch M, ‘AI and the future of mediation’ Issue 30, Conciliation Resources, February 2024 https://share.google/X0euu3gCJd03linVY on 1 October 2025.
[25] The Role of AI in Alternative Dispute Resolution: Mediation and Arbitration,” Mediate, March 14, 2025.
[26] The Role of AI in Alternative Dispute Resolution: Mediation and Arbitration,” Mediate, March 14, 2025.
[27] Wäehlisch M, ‘AI and the future of mediation’ Issue 30, Conciliation Resources, February 2024 https://share.google/X0euu3gCJd03linVY on 1 October 2025.
[28] Wäehlisch M, ‘AI and the future of mediation’ Issue 30, Conciliation Resources, February 2024 https://share.google/X0euu3gCJd03linVY on 2 October 2025.
[29] Vyas K, ‘Human-in-the-Loop: Maintaining Control in an AI-Powered World’ (2025) https://share.google/MEJhBmGcIM3AhV4MM on 1 October 2025.
[30] Vyas K, ‘Human-in-the-Loop: Maintaining Control in an AI-Powered World’ (2025) https://share.google/MEJhBmGcIM3AhV4MM on 1 October 2025.
[31] Vyas K, ‘Human-in-the-Loop: Maintaining Control in an AI-Powered World’ (2025) https://share.google/MEJhBmGcIM3AhV4MM on 1 October 2025.
[32] Vyas K, ‘Human-in-the-Loop: Maintaining Control in an AI-Powered World’ (2025) #https://share.google/MEJhBmGcIM3AhV4MM on 1 October 2025.
[33]Giacalone M, ‘AI and the future of private dispute resolution mechanisms’ in Schmitz A J, Ortolani P and Giacalone M (eds), The Cambridge Handbook on AI and Civil Dispute Resolution (Cambridge University Press forthcoming 2025),15-16.
[34]Giacalone M, ‘AI and the future of private dispute resolution mechanisms’ in Schmitz A J, Ortolani P and Giacalone M (eds), The Cambridge Handbook on AI and Civil Dispute Resolution (Cambridge University Press forthcoming 2025),15-16.
[35]Fasanya C F, ‘Using Artificial Intelligence (AI) in ADR: Opportunities and Challenges’ The University of America, Curaçao, 2023 – https://share.google/gGRaMjaprM9siTFud on 1 October 2025
[36]Fasanya C F, ‘Using Artificial Intelligence (AI) in ADR: Opportunities and Challenges’ The University of America, Curaçao, 2023 – https://share.google/gGRaMjaprM9siTFud on 1 October 2025
[37] Judiciary of Kenya, Judiciary Artificial Intelligence Adoption framework: aligned with social transformation through access to justice, 11 August 2025, page 28.
[38] Judiciary of Kenya, Judiciary Artificial Intelligence Adoption framework: aligned with social transformation through access to justice, 11 August 2025, page 28.
[39] Judiciary of Kenya, Judiciary Artificial Intelligence Adoption framework: aligned with social transformation through access to justice, 11 August 2025, page 28.
[40] Judiciary of Kenya, Judiciary Artificial Intelligence Adoption framework: aligned with social transformation through access to justice, 11 August 2025, page 28.
[41] Article 159, Constitution of Kenya, 2010.
