AI in Dispute Resolution: A Practitioner's Guide
- Shiv Martin
- Jul 7
- 8 min read
A lot of the conversation about AI in dispute resolution is happening in conference panels, academic papers and global vendor demonstrations. The conversation that is not yet happening as much is the one between practitioners about what AI is actually changing in the room.
What I have noticed over the past two years is that AI has stopped being a future-tense topic for mediators, conciliators and tribunal members. It is in the room with us now. Parties arrive with AI-drafted opening statements. Lawyers bring AI-summarised chronologies. Self-represented parties ask AI to interpret what the mediator just said, sometimes during a break, sometimes in real time. Tribunal applications include AI-generated arguments and occasionally AI-invented case citations.
This is not a guide to AI as technology. There are plenty of those, and most of them are written by people who do not sit across from parties in conflict for a living. This is a guide to AI as practice. What is changing, what is not, and what skills practitioners need to bring into the next mediation, conciliation or tribunal hearing they walk into.
AI has stopped being a future-tense topic for mediators, conciliators and tribunal members. It is in the room with us now.
What is actually changing in the dispute resolution room?
Three shifts are worth naming, because they are the ones that show up consistently across the work, regardless of whether the matter is workplace mediation, commercial conciliation, tribunal review or community dispute.
The first shift is the preparation gap. Parties used to arrive at mediation having had limited contact with the issues since they wrote their statement, often weeks earlier. They now arrive with AI-assisted summaries of every email exchange, every previous statement, every relevant section of the contract or policy. They are more prepared on the documents than mediators sometimes are. Whether their preparation is accurate is a separate question.
The second shift is the language escalation. Parties using AI tend to arrive with submissions that are more legally formal, more confident in tone, and more polished than the underlying situation warrants. The communication style runs ahead of the actual dispute. A workplace disagreement that would once have been described in plain language now arrives written like a tort claim. The risk is not the polish itself. The risk is what gets lost between the plain language version and the polished one, including the emotional truth of what is actually wrong.
The third shift is the real-time second opinion. Parties now ask AI tools to validate the mediator's process choices, suggest counter-positions during breaks, or interpret what the other party just said. Sometimes this helps the party feel less alone in the room. Sometimes it works against the slow, careful pace that mediation depends on. Often it does both at once.
What is not changing, and why that matters?
The work itself is not changing. The reason parties come to mediation, conciliation or tribunal is still that they have a dispute they cannot resolve on their own, and they need a structured process and an independent practitioner to help them think it through.
Parties still arrive carrying history, expectations, fear, loss and meaning-making. They still need to feel heard before they can hear. The signal of a mediation working is still the moment a party stops repeating their position and starts asking a different question.
What is not changing is also the practitioner's role. Mediators are still process custodians. Conciliators are still independent third parties helping a structured process unfold. Tribunal members are still applying legal standards to contested facts. AI is not replacing any of that work in the rooms I sit in, and I do not expect it to in the rooms most readers of this article sit in either.
The reason this matters is that the practitioner who anchors in what is not changing is in a much better position to respond well to what is. If the work itself is still the work, then AI is a context shift the practitioner navigates, not an existential threat to the practice. That framing matters for how we approach the rest of this guide.
The practitioner skills that matter now
There is a tendency in the current AI-in-dispute-resolution conversation to assume that practitioners need new skills they do not have. In my experience, the skills that matter most are the ones experienced practitioners already use, applied with more deliberateness because the room is now louder, faster and more layered than it was three years ago.
Five skills are worth naming.
Reality-testing AI-influenced material. When a party arrives with an AI-drafted statement that reads as legally sophisticated, the practitioner's job is the same as it has always been: to test whether the polished version reflects the actual situation. Useful questions are not adversarial. "Tell me in your own words what happened first" is a question that works whether the statement was drafted by AI, by a lawyer, or by the party themselves. The discipline is to ask it.
Pacing against the AI-amplified speed. Parties using AI in real time can move faster than the process is built to absorb. The practitioner skill is not to match that speed. It is to maintain the slower pace that mediation depends on, even when one party is asking AI for a counter-position every five minutes during a break. The structure of the process is the practitioner's responsibility, not the parties'.
Naming AI use without shaming it. If a party has used AI to prepare, the practitioner does not need to pretend they have not. A simple acknowledgment ("I can see this submission was carefully prepared, possibly with some AI assistance") names what is in the room and gives both parties permission to talk about how their material was put together. Naming reduces tension. Pretending not to notice does the opposite.
Distinguishing AI assistance from AI dependence. A party using AI to draft a statement they have read, edited and stand behind is different from a party using AI to argue a position they do not fully understand. The practitioner skill is to notice the difference, often through a single clarifying question. The first situation is fine. The second is a procedural fairness concern worth thinking about.
Holding the human centre. What experienced mediators know is that the parties' actual interests rarely live in the language of the submissions. They live in what the parties say when the submissions are set aside and the practitioner asks a different kind of question. The skill is not to be distracted by the polish of AI-drafted material into thinking that material is what the parties actually need to talk about.
AI in process versus AI by parties (a useful distinction)
One of the most common conflations in the current conversation is the difference between AI being used to operate the dispute resolution process itself (a tribunal using AI to triage cases, a platform using AI to suggest settlements, an online dispute resolution tool that asks adaptive questions) and AI being used by the parties bringing the dispute (a complainant drafting their submission with AI, a respondent using AI to summarise the case file, a lawyer running their opening statement through an AI for tightening).
These are different conversations.
AI in the dispute resolution process raises questions about institutional design, transparency, accountability and whether automated decision-making is appropriate for the kind of matter being heard. Those are important questions, but they are mostly being answered at the policy and platform level, not by individual practitioners.
AI by parties is a practitioner-level question. It is the one that shows up in the next mediation a reader of this article walks into. The skills above are about AI by parties, because that is the conversation practitioners can actually do something about.
The distinction matters because the policy debate about AI in dispute resolution processes is sometimes used as a reason for practitioners to feel that the field is at an unsettling inflection point. It may be at an inflection point at the institutional level. At the practitioner level, the work is more continuous than the headlines suggest.
What this means for procedural fairness
Procedural fairness is the legal and ethical backbone of most Australian dispute resolution practice. It is also where AI is creating the most genuinely new questions for practitioners, particularly for those working in tribunal or quasi-judicial settings.
The traditional procedural fairness checks were built for a context where parties wrote their own submissions and responded to each other's positions in real time. AI assistance complicates two of those checks.
The first complication is the opportunity to be heard. If one party has used AI to prepare a sophisticated submission and the other party has not, is the second party still meaningfully heard, or has the playing field tilted in a way that the practitioner needs to manage? In my view, the practitioner's role is not to police AI use. It is to make sure both parties have a fair opportunity to put their case, which may sometimes mean offering more time, more clarification questions, or more support to the party who has not used AI.
The second complication is the right to know the case against you. If a submission contains AI-generated arguments or AI-invented citations, the party responding has a fair-process concern, particularly if they cannot easily verify what is genuine and what is fabricated. The practitioner's role is to flag this in the process rather than ignore it. A simple direction ("any submission relying on legal authority must include a verifiable citation") is often enough.
A separate article on this site goes into procedural fairness in AI-assisted complaints in more depth. It is worth reading for practitioners working in regulatory or Ombudsman contexts. For tribunal members, conciliators and mediators working with party-driven AI use, the two complications above are the practical ones to keep in mind.
A note on AI training for dispute resolution practitioners
The training conversation for practitioners is, in my experience, more useful when it is focused on practice than when it is focused on technology. Most mediators, conciliators and tribunal members do not need to understand the architecture of a large language model. They need to know how to recognise an AI-influenced submission, how to ask the clarifying questions that get to the underlying interests, and how to maintain the procedural fairness checks the room depends on.
When organisations ask me about training, my recommendation is usually that they prioritise the practitioner skills over the technical knowledge. A two-hour session on the five skills above does more for an experienced practitioner than a full day on how generative AI works. The exception is for practitioners who are designing or implementing AI tools in their own practice, which is a smaller and more specialised audience.
Free guide: Responding to AI-Generated Complaints
A practical intake tool for complaint handlers, conciliators and dispute resolution practitioners working through AI-assisted correspondence. Focus on substance, ask better questions, and progress matters fairly. Download the guide here.
What I have learned from sitting in mediation rooms for fifteen years is that the technology in the room changes more often than the work itself does. Email changed how parties communicated before the mediation. Video conferencing changed where the mediation happened. AI is changing how parties prepare and how some of them participate.
The practitioner who anchors in their own role, maintains the pace of the process, and holds the human centre of the work is in the same position they have always been in. Better prepared parties, faster-moving parties, AI-influenced submissions. These are conditions to be worked with, not threats to be defeated.
If your team is thinking through what AI means for your dispute resolution practice, I am happy to have a conversation about it. The practitioner skills above are the starting point. The deeper work (policy, system design, in-house training) comes after the practitioner foundations are clear.
Hi, I'm Shiv Martin. I'm a nationally accredited mediator, lawyer, conciliator, and conflict management specialist with over a decade of experience working across government, business, and community settings. I support teams to navigate complex and emotionally charged situations through mediation and conciliation, conflict skills training, facilitation, and practical advice on policies and processes. My approach is grounded in law, psychology, and real-world dispute resolution, with a strong focus on clarity, fairness, and workable outcomes.
If you'd like to talk about how I can help you or your organisation, you can get in touch here: 👉 Contact us
Comments