Find Mediators Near You:

Designing Multi-Agent, Multi-Outcome Dispute Systems

Parties in complex disputes often send lawyers with authority as agents to negotiate on their behalf. Of course, these agents often come back with a single negotiated solution, if any agreement at all. Consider if parties on both sides had the resources to send twenty lawyers each as agents on their behalf. Every lawyer would have authority to negotiate their own outcome with a lawyer from the other side. Instead of one negotiated outcome, twenty possible solutions could be returned to each party—each outcome independent, and some more enticing and creative than others. 

I envision a creative new dispute resolution model that could achieve this multiple-agent system and enable multiple outcomes. Parties could then see a range of viable—and satisfactory—outcomes to inform further negotiation, or even agree on one outcome to implement. 

The Need for Advanced Dispute Systems

In today’s complex world, negotiation and mediation stretches far beyond simple disputes and enters the realm of highly intricate conflicts, whether they be personal, corporate, or international. Traditionally, evaluative dispute resolution processes through arbitration or courts focus heavily on singular win-lose outcomes, often overlooking the deeper interests and potential creative solutions that could benefit all parties involved. Realizing the limitations of current systems, I suggest a more innovative approach to resolving conflicts that not only broadens the spectrum of outcomes but also tests various strategies to enhance satisfaction among disputants.

Dispute System Design (DSD) has emerged as a strategic framework that tailors conflict resolution mechanisms to the specific needs of the parties involved. A well-crafted dispute system evaluates fairness, efficiency, satisfaction, and the recurrence of disputes, among other factors. By prioritizing these criteria, the system aims to resolve conflicts in a manner that parties find just and equitable, potentially reducing future disputes and fostering a positive long-term relationship. The more solutions that meet the parties’ needs, especially through creative means, the more likely these benefits will be attained. 

In resolving disputes, distinguishing between interest-based and power-based processes is vital. While power-based methods hinge on legal rights and authoritative stances, interest-based processes (such as mediation, and often, effective negotiation) delve into the underlying needs and values of the parties, fostering more creative and mutually satisfactory outcomes. Such processes are especially beneficial in complex disputes involving many variables or many parties, where rigid adherence to rights or power might obstruct potential integrative solutions.

Mediation is a powerful tool for navigating through negotiation deadlocks. By focusing on interests rather than positions, mediators help parties uncover shared values and explore innovative solutions. This approach is particularly useful in multifaceted disputes that may otherwise end in stalemate. Additionally, mediators can act somewhat like an agent for each party when reframing their perspectives and caucusing between parties. However, a single mediator can only do so much to get parties to move past impasse and towards a single possible solution. 

Introducing the Multi-Agent, Multi-Outcome Mediation Model

Recognizing the limitations of traditional dispute systems—particularly the difficulties in getting parties to have open dialogue and brainstorm creative solutions, even during interest-based processes like mediation—I propose a model that utilizes multiple agents to represent the parties in several parallel negotiation sessions. This model allows each agent to negotiate independently with an agent of the other party based on predetermined parameters, fostering a range of potential outcomes that might not surface in a single-threaded negotiation. In other words, I envision a process akin to a party giving twenty lawyers authority to negotiate independently with twenty lawyers from the opposing party, and each lawyer to return to the party with a discrete, independent outcome from their negotiations, which the party can then approve or reject. 

Key Features of the Multi-Agent Model

  1. 1. Diversity of Strategies: By employing multiple agents, the model allows a party principal to test various negotiation strategies, enhancing the likelihood of identifying solutions that satisfy all parties.
  2. 2. Enhanced Creativity: The independence of each agent encourages creative thinking, free from groupthink or other constraints of a collective negotiation session, while still allowing for the diversity of perspectives which can bolster creative solutions.
  3. 3. Range of Outcomes: The model produces a spectrum of potential viable outcomes, providing parties with a broader understanding of what might be acceptable and achievable.
  4. 4. Approval or Rejection of Outcomes: Party principals can approve or reject outcomes which agents return, and outcomes which both party principals accept can form the basis of an agreement, or at the very least further nuanced negotiation. 
  5. 5. Confidentiality: The process would need to ensure confidentiality to the parties so that data, systems, and human components are secured against disclosure, allowing the benefits of candor by the parties in giving negotiating power and instructions to their agents. 
  6. 6. Anonymity of Opposing Party’s Agent Strategies: For the model to work effectively, parties cannot be privy to the actual events in the agent-to-agent negotiations (such as the opposing side’s strategies employed by different agents), or else this would defeat the purpose of obtaining independence in each interaction, and possibly cause the parties to restrict their candor and the range of strategies they would test. 

Potential for Artificial Intelligence Integration

Looking ahead, integrating Artificial Intelligence (AI) into this model could significantly enhance its efficiency and scalability. AI agents, programmed with the ability to learn and adapt, could handle negotiations, test strategies, and generate outcomes at a scale unattainable by human agents alone. Where ten or forty human negotiator agents would need to negotiate over substantial time periods, AI could potentially run dozens or hundreds of interactions in short order, testing strategies and developing outcomes. Such technological advancement could revolutionize dispute resolution, making it more accessible and less costly. Parties could then review and approve or disapprove the AI-generated outcomes. The challenges going forward are to test this model and further develop AI processing, computer learning, and ability to come up with creative solutions that can benefit the parties. 

Conclusion: A Path for Future Dispute Resolution

The multi-agent, multi-outcome approach represents a paradigm shift in how we handle complex disputes. By embracing this model, we can transform conflict resolution for complex disputes into a more dynamic, integrative, and satisfying process. Refining this model, both through human-mediated experiments and AI integration, has the potential to create a universally applicable system where dispute resolution is not just about finding a single compromise but about crafting a spectrum of outcomes that offer effective and creative resolutions.

For more information on this model, or for a copy of my full paper fleshing out this proposed systems design, feel free to contact me at [email protected]


Keith Richards

Keith Richards is a Staff Editor for OSJDR. Keith graduated from the University of Texas at Austin with a Bachelors in Psychology and French and a Business Foundations Certificate, and graduated from Ball State University with a Masters in Psychology. Keith serves in the U.S. Air Force and will be… MORE >

Featured Members

View all

Read these next


Beyond Adjudication: Resolving International Resource Disputes in an Era of Climate Change

I. INTRODUCTION Climate change is one of the greatest emerging threats to global peace and security. Among other impacts, climate change will exacerbate the scarcity of water, food, and other...

By Anna Spain

Why Certify Public Policy Mediators?

From John Folk-Williams's blog Cross Collaborate Image Credit: Arcady – In his keynote address at the 2009 conference of the Association for Conflict Resolution, Wallace Warfield discussed the difficulty...

By John Folk-Williams

Customizing The Mediation Process

From Michael P. Carbone’s Mediation Strategies Blog Back on September 1, I wrote about the “Task Force on Improving Mediation Quality.” The Task Force, which was formed by the American...

By Michael P. Carbone