Artificial Intelligence (AI) in Arbitration

Artificial Intelligence (AI) is now ubiquitous in our everyday lives. The CSIRO describes AI as “maths, not magic”. It is technology that uses algorithms to imitate human thinking, our problem-solving abilities, and includes machine learning. Machine learning allows computers to “automatically” learn and develop, and deep learning models can create original content through the use of algorithm models to “see” patterns in data to make predictions.

With the expansion of AI-assisted technologies, there are opportunities for AI to assist in the facilitating and conducting of arbitrations. While taking advantage of these opportunities, we also need to be alert for issues with using AI in arbitration.

CIArb has prepared a guideline on the use of AI in arbitration:
https://www.ciarb.org/media/bpndtcgu/guideline-on-the-use-of-ai-in-arbitration_updated-sept-2025.pdf
This guideline also discusses the benefits and risks of using AI.

Taking this into account, discuss:

1. Whether AI can be used to automate certain types of arbitration or parts of the arbitration process, to make it more accessible to a wider group by making it more affordable, what are the risks in doing so, and what mechanisms could be adopted to mitigate these risks; and
2. Whether the enforceability of an arbitral award can be challenged if AI is used during the arbitration, or by the arbitrator, and what mechanisms could be adopted to mitigate these risks.

You may wish to consider:

• Using AI to predict award outcomes
• Using AI to predict opposition submissions
• Using AI to predict enforcement issues in various jurisdictions (and how sympathetic local courts will be)
• Using AI to predict an arbitrator’s likely bias
• Ethical and bias considerations in AI