By Chandni Gopal, Partner, Prianka Soni, Senior Associate & Amaarah Mayet, Associate, Webber Wentzel

As the "great disruptor" of our time, artificial intelligence (AI), with its inherent ability to perceive, reason and solve problems, has enormous capacity to shape international arbitration practice.

The use of AI to enhance technological efficiency has long been accepted by the international arbitration community. In recent times, AI has emerged as a useful instrument that, when applied properly, can both increase the acceptance of international arbitration as the preferred forum for resolving commercial dispute resolution.

Current uses of AI

In international arbitration, the use of stenographers, short-hand techniques and backup audio recordings for the manual transcription of proceedings are not-too-distant memories. The need for time, resource and cost efficiencies propelled rapid technical breakthroughs that encouraged investment in machine learning and natural language processing (NLP) AI technologies. A direct offspring of these early technologies, voice-to-text technology, revolutionised transcription services, resulting in improved accuracy in record time, at a fraction of the cost.

It did not take long for machine learning, NLP and generative AI to be deployed in making document-intensive stages of the arbitration process considerably more efficient. Electronic discovery programmes, for example, can analyse huge amounts of electronically stored information, identify relevant material, and automatically collate them for discovery in a matter of hours (instead of days). In the discovery context, multilingual document review and translation software can offer initial content interpretation without the need for human linguists.

The international arbitration community has demonstrated sophistication in its receptiveness to new technologies while acknowledging the importance of appropriate checks and balances in protecting the integrity of the arbitration process. Translation software, for instance, may assist parties in identifying potentially relevant material despite it being in another language, but procedural rules generally afford the tribunal the discretion to direct the manner and form of translating documents for reliance purposes.

The international arbitration community has undoubtedly opened up a world of possibilities with both exciting prospects and challenging situations due to its controlled approach to AI-powered technologies.

Potential uses of AI

By harnessing AI technologies, arbitrators and legal practitioners can streamline the case management process. Firstly, AI can automate tedious tasks like analysing vast amounts of documents. Contracts, legal precedents, and evidence can be scanned in a fraction of the time it takes humans, allowing legal professionals to redirect their focus. Instead of sifting through mountains of paper, they can concentrate on building strong legal arguments and crafting persuasive submissions for their clients.

AI can also predict potential outcomes or suggest optimal dispute resolution strategies. By analysing past arbitration awards, relevant legal decisions, and arbitrator track records, AI algorithms can empower parties involved, including third-party funders, to make informed settlement decisions and the likelihood of claim success.

Witness preparation can also be streamlined with the help of AI-powered tools. These tools can analyse case documents, transcripts, arbitrator awards, and legal precedents to identify key issues and themes. This allows legal teams to create clear and persuasive submissions, develop tailored questioning strategies, highlight potential strengths and weaknesses in witness testimony, and ensure consistency in the narrative presented during hearings. Furthermore, AI can simulate cross-examination scenarios, helping witnesses anticipate tough questions and prepare effective responses.

Arbitrators themselves can benefit from AI's supportive capabilities. Machine learning AI can offer valuable assistance by creating preliminary timetables based on procedural rules, evidence rules, and party availability. AI can also assist with drafting routine sections of awards, such as factual backgrounds and historical context. By automating repetitive tasks, AI frees up arbitrators' time to focus on complex legal analysis and decision-making. Additionally, AI algorithms can play a vital role in quality control. By comparing language patterns, legal reasoning, and factual assertions in draft awards against established precedents and party submissions, AI can flag inconsistencies, factual errors, or potential biases. This not only enhances the quality and reliability of awards but also safeguards the integrity and fairness of the entire arbitration process.

The concept of a fully automated AI arbitrator remains a topic of debate. Concerns about AI hallucinations and biased reasoning are well-founded. However, AI's supportive capabilities offer intriguing possibilities. In cases with standardised contracts, clearly defined decision-making frameworks, or settled legal principles, AI arbitrators could expedite proceedings, reduce costs, and deliver more predictable outcomes. Even with the potential benefits, concerns about transparency, fairness, and adherence to the rule of law must be carefully addressed to ensure the legitimacy and acceptance of AI-driven international arbitration.

Responsibly Embracing Change

In essence, AI's role in arbitration goes beyond mere efficiency gains. It has the potential to fundamentally transform the nature of dispute resolution by augmenting arbitrators' capabilities, enhancing decision quality, and reshaping the dynamics of legal proceedings.

As AI becomes more sophisticated and integrated into the legal profession, stakeholders must navigate the complexities and opportunities presented by this paradigm shift. The rapid development of AI technologies underscores the urgency for the international arbitration community to adapt responsibly. This includes complying with new regulations and guidelines are ensuring professionals are properly trained.

The Silicon Valley Arbitration and Mediation Center, for example, has already released draft Guidelines on the Use of Artificial Intelligence (AI) in International Arbitration for Public Consultation (SVAMC Guidelines). The Guidelines place an emphasis on understanding AI’s limitations and risks, safeguarding the confidentiality, ensuring competent and diligent use of AI (including appropriate disclosure of its use), and ensuring that arbitrators do not delegate their decision-making responsibilities. The SVAMC Guidelines indicate that AI is becoming more involved in arbitration, and they serve as a useful reference guide for using AI's potential in arbitration safely and responsibly. International arbitration institutions are likely to follow suit by releasing their own guidelines in compliance with applicable laws and regulations.

The proliferation of AI regulation worldwide will also create peculiar challenges in the international arbitration community. AI technology used in international arbitration will need to remain compliant with new and differing regulations across various relevant jurisdictions. For instance, it is quite possible that different jurisdictions with different AI regulatory regimes house the laws governing the content of a dispute, the procedure of a dispute, and the venue at which the arbitration proceedings are held as agreed between the parties in the arbitration clause and agreement. In addition, it is important to the enforceability of the arbitral award that the award is rooted in the mandatory laws of the place where the award is rendered and the place where the enforcement of the award is sought.

These challenges also present an opportunity for international arbitration. When there are differences and conflicts in applicable laws, it creates an opportunity for practitioners to be creative. Thus creativity could lead to the harmonisation of various laws, or contribute to the development of flexible AI systems which are compliant and fit for purpose.

Overall, the use of AI in international arbitration requires a change-positive approach characterised by agility, adaptability and flexibility, as well as principled decision-making in which the rule of law, underlying purpose and legitimacy of international arbitration are safeguarded.