By Eric Gillett
I have adult children, and they understand artificial intelligence, “AI.” I have a very limited understanding of AI, which, according to my kids, is consistent with other adults in their 60s. Obviously, I didn’t grow up with AI. I didn’t even grow up with access to a computer until I was in college. So, it’s unsurprising to my kids that I lack an intuitive understanding of what AI is or what it can accomplish. Instead, I have to listen to them and ask questions because, honestly, I’m still amazed by a Google search.
AI has found its way into the legal profession and not just as a way to write a brief. For younger lawyers, it is becoming the way they process thoughts. You have probably heard of “big data.” To me, that means a lot of information that can be reviewed by a computer. When you apply AI to big data, you blow the doors off the way you can look at all that data. You can ask AI to evaluate contracts, compare documents, and look for inconsistencies between documents. You can ask AI to examine a plaintiff’s medical records and compare them to deposition testimony, not just testimony of the plaintiff but also of her or his doctors. You can ask AI to evaluate the risk of a liability finding and estimate what amount of damages can be expected based on a history of verdicts or settlements in your jurisdiction.
Whatever you can imagine as a task for the smartest lawyer, you can ask AI to take it on and do it quicker. And we’re just getting started. I’m having a hard time imagining where AI will be even in two years. Some of what I imagine is exciting, some of it is worrisome.
As AI continues to reshape parts of the legal profession, it’s making its way into mediation. Some of the promised benefits are compelling: efficiency, data-driven insights, and streamlined processes. But as with any new tool — especially one that touches on sensitive disputes — we need to take a closer look. Not just at what AI can do, but where it might cause problems.
Below are five key concerns that we should be thinking about when AI enters the mediation space. Whether you’re actively using these tools or just beginning to hear about them, these are the issues that will shape how AI fits — or doesn’t — into the future of dispute resolution.
1. Bias and Fairness:
The Algorithm Isn’t Neutral
AI systems rely on historical data, which means they can replicate the biases embedded in that data. If past outcomes reflect racial, gender, or socioeconomic disparities, the algorithm doesn’t correct for that — it reinforces it.
In mediation, where fairness and trust are paramount, this is a real problem. Compounding that is the opacity of many AI tools. When you can’t explain how a recommendation was generated — whether it’s a suggested settlement figure or a case strength analysis — you can’t vet it for fairness, and parties are right to be skeptical.
That’s why human oversight is essential. AI can assist, but it cannot replace the nuanced judgment of a skilled mediator or counsel who understands the context, the people, and the power dynamics in play.
2. Confidentiality and Data Security: Core Principles at Risk
Confidentiality is foundational to mediation, but AI complicates that. Tools that process or store mediation data could inadvertently expose sensitive information — especially if that data is later used to train AI models. Even anonymized data can carry re-identification risks.
There’s also the cybersecurity layer. AI platforms are attractive targets for data breaches. If mediation materials are stored in the cloud or transmitted through third-party tools, lawyers need to understand the safeguards in place.
If you’re involved in a mediation using AI, ask the hard questions: Where is the data stored? Who has access? Is there a clear, written consent process? Data security isn’t a technical detail — it’s a legal and ethical obligation. Various jurisdictions are beginning to issue ethical opinions on AI. But the importance of managing confidentiality cannot be overstated.
3. Ethical and Professional Obligations: Where AI
Meets Legal Duties
The introduction of AI into mediation raises important ethical issues. Transparency is essential. Parties should be informed if AI tools are being used, what their purpose is, and how they might influence the process or outcomes.
Then there’s the question of control. AI can offer recommendations or data visualizations, but it cannot — and should not — replace the judgment of the mediator or legal counsel. We are bound by ethical duties that require independent judgment and reasoned decision-making.
And what about accountability? If an AI-assisted process leads to an unjust or inaccurate outcome, who’s responsible? The developer? The platform provider? The mediator? As things stand, the liability framework is still evolving, and until it catches up, practitioners need to tread carefully.
4. Built-in Limitations:
What AI Can’t Do (Yet)
AI has analytical power, but it lacks emotional intelligence. It can’t read a room, manage interpersonal dynamics, or recognize the subtle non-verbal cues that often shape a mediation’s trajectory. As I’ve discussed in many other articles, your mediator’s ability to build trust is driven by his or her ability to project emotional intelligence.
Cultural awareness is another limitation. AI tools are often developed with one dominant cultural context in mind and may miss or misinterpret differences in language, communication styles, or social expectations. In our multicultural society, cultural awareness is essential.
Most importantly, AI lacks the ability to think creatively. Mediators regularly help parties find inventive, customized solutions. AI works within programmed boundaries — it doesn’t improvise (yet). That limitation matters in a process that relies heavily on adaptability and lateral thinking.
5. Practical Challenges: Training, Trust, and Integration
Even with the best intentions, integrating AI into mediation presents logistical challenges. Most mediators and attorneys haven’t been trained to work with AI tools. Understanding how these tools function — and how to explain them to clients — will require a learning curve.
As discussed above, there’s also the issue of trust. Some parties (and practitioners) may resist AI because it feels impersonal or overly technical. Mediation is a human-centered process, and there’s understandable concern about technology shifting the tone or balance of the discussion.
Finally, implementation isn’t plug-and-play. AI tools must be carefully integrated into existing workflows, accounting for process changes, ethical review, and cost considerations. Without that, the technology risks becoming more of a distraction than a benefit.
Moving Forward: A Balanced, Responsible Approach
AI isn’t going away. But neither are the foundational principles of mediation: fairness, confidentiality, human judgment, and dignity. As the technology evolves, we have a responsibility to ensure its use reinforces — rather than undermines — those values.
AI only became popular a couple of years ago and has gained popularity at a phenomenal rate. Futurists speculate that AI will begin thinking for itself in five years or less. What that means for mediation in particular and the legal profession in general is hard to say. But it is here and being used by our legal profession more and more. So it is imperative that we work with it.
Used wisely, AI has the opportunity to enhance mediation. But we need to proceed with caution and intentionality. After all, the people involved in mediation, not AI, are the point.
Eric Gillett is a founding member and managing partner at Preg, O’Donnell & Gillett. He is licensed to practice in Washington, Oregon, and Alaska and has tried dozens of cases to verdict and mediated hundreds more. A highly experienced commercial mediator, Eric can be reached through his legal assistant, Jasmine Reddy, at 206-287-1775 or jreddy@pregodonnell.com. Further information is available at www.gillettmediation.com or via email at eric@gillettmediation.com or egillett@pregodonnell.com.
