Artificial intelligence is no longer a speculative topic in mediation; it is already reshaping how parties and mediators prepare for disputes, how mediators manage complexity, and how parties experience the process itself. In a recent discussion featuring Gary Doernhoefer, founder of ADR Notable , the conversation moved quickly past whether AI belongs in mediation and focused instead on how it can be integrated in ways that strengthen the core values of the field.
What emerges from this conversation, and from Gary’s broader work, is a clear through‑line: with current technology, AI’s greatest potential in mediation is as a coach and preparation partner, not as a decision-maker or substitute for professional judgment. As the field looks ahead, preparation of mediators and of parties may be where AI has its most immediate and profound impact.
AI as Cognitive and Process Coach
One of the most promising roles for AI in mediation is as a coach that supports human decision‑making, reflection, and readiness. Mediation is cognitively demanding: mediators track facts, emotions, interests, power dynamics, and procedural choices, often simultaneously. Parties, meanwhile, arrive with incomplete information, unrealistic expectations, or little understanding of what mediation will actually require of them.
AI tools are well‑suited to support both sides of this equation.
Rather than directing outcomes, AI can:
- Help mediators
- Reflect on process choices
- Develop background on complex issues
- Prompt consideration of overlooked issues or interests
- Assist parties
- Clarify and develop goals, priorities, and alternatives
- Normalize emotional and strategic preparation for difficult conversations
- Workshop potential strategies for mediation
- Reflect on the consequences of failing to settle
This “coaching” function mirrors what experienced mentors and trainers have long provided, but in a way that is scalable, timely, and responsive in real time. As Doernhoefer emphasizes, the goal is augmentation, not substitution: AI helps mediators and parties show up more prepared, more informed, and more intentional.
Preparation: Where AI May Matter Most
Decades of mediation research and practice converge on a simple insight: prepared participants settle more often and more effectively. Scholars such as Rachael Wissler have shown that when parties understand the mediation process, have realistic expectations, and arrive ready to engage substantively, the likelihood of settlement increases significantly.[1] Preparation shapes not only outcomes, but party satisfaction and perceptions of procedural justice.
AI opens new possibilities for delivering this preparation at scale.
For parties, AI‑driven preparation tools can:
- Explain the mediation process in plain language
- Help users articulate interests, concerns, and priorities before the session
- Prompt reflection on BATNAs, risks, and tradeoffs
- Reduce anxiety by demystifying what will happen and what is expected
For mediators, these tools can surface patterns and readiness gaps before the session begins, allowing the mediator to intervene proactively—by adjusting process design, expectations, or pacing. Doernhoefer points out that AI provides dynamic responses to simple (e.g., how should the mediator be addressed?) and complex (e.g., if the mediator uses bracketing, how should that shift my thinking about my case, and the other party’s perspective). In this sense, AI becomes a front‑end enhancement to mediation quality, improving the conditions under which mediation takes place rather than attempting to influence the outcome itself.
AI and Coaching for Mediators: Building on Existing Models
The idea of AI as a coach fits naturally with existing efforts to professionalize and support mediation practice. John Lande’s work with the RPS (Reflective Practice System)[2], for example, has long emphasized structured self‑reflection, feedback, and continuous improvement for mediators. RPS helps mediators think critically about their choices, habits, and assumptions—core elements of professional growth.
AI can extend and reinforce this kind of reflective practice by:
- Prompting post‑session reflection questions
- Identifying patterns across cases
- Highlighting recurring challenges or blind spots
- Offering process‑focused suggestions grounded in best practices
Importantly, this kind of coaching does not threaten mediator autonomy. Instead, it supports what effective mediators already do: think deliberately about how their actions shape the process. AI simply makes that reflection more accessible and consistent.
Transparency, Ethics, and Party Trust
A recurring concern around AI in mediation is the fear of opaque “black box” systems making decisions that affect party outcomes. The coaching and preparation model avoids much of this risk. When AI is used to support readiness, reflection, and understanding, rather than to recommend specific settlement terms or outcomes, it remains aligned with mediation’s ethical commitments to self‑determination, neutrality, and informed consent.
ADR Notable’s development philosophy reflects this orientation. AI‑enabled tools are designed to be visible, optional, and subordinate to professional judgment. They support mediators’ thinking without dictating their choices and support parties’ preparation without steering their substantive decisions.
This distinction matters. AI that prepares and coaches strengthens party autonomy; AI that decides undermines it.
On the Horizon: AI Literacy as a Core Mediation Skill
Looking ahead, the next horizon for AI in mediation may be cultural rather than technical. As preparation and coaching tools become more common, AI literacy will increasingly become part of professional competence for mediators—much like ethics training or continuing education.
Doernhoefer urges mediators to learn whether and how to use AI. He is not concerned that bots will take over the business of mediation in the near future, but does worry that individuals who do not have experience mediating will use bots to replace traditional mediators. He points out that bots are already replacing customer service agents, but are not yet able to navigate complex, emotional, and human disputes. For example, he asks how, beyond linguistics, a bot will be able to recognize the value, meaning, timing, and weight of an apology in a mediation session.
This does not mean mediators must become technologists. It does mean understanding:
- How AI is being used in their practice
- What data is collected and protected
- How AI outputs should (and should not) inform human judgment
The mediators who thrive in this next phase will be those who see AI not as a threat, but as a structured way to reinforce what mediation already values most: thoughtful preparation, reflective practice, and humane engagement with conflict.
Conclusion: From Tool to Trusted Coach
The future of AI in mediation will not be defined by dramatic replacements of human expertise, but by quieter, more consequential shifts in how people prepare for difficult conversations. If deployed thoughtfully, AI can act as a coach in the background, helping mediators be more intentional and helping parties arrive ready to engage constructively.
As the discussion with Gary Doernhoefer and the work of ADR Notable suggest, the real promise of AI lies upstream before the session begins and after it ends, where preparation and reflection shape everything that follows. In that role, AI does not change what mediation is. It helps mediation be what it has always aspired to be, more often and for more people.
[1] See Roselle L. Wissler & Art Hinshaw, What Happens Before the First Mediation Session? An Empirical Study of Pre-Session Communications, 23 Cardozo Journal of Conflict Resolution 143 (2022).
[2] See John Lande, A Practical Guide for Using the RPS Negotiation and Mediation Coach (2025), https://scholarship.law.missouri.edu/facpubs/1264.
See below for the conversation with Gary Doernhoefer of ADR Notable.
