Efficiency vs. fairness

Abubaker Karmustaji examines DIAC’s 2025 partnership with Jus Mundi, exploring how emerging AI research tools intersect with evidentiary integrity, procedural fairness and the practical pressures of expedited proceedings under Article 32.

In June 2025, the Dubai International Arbitration Centre (DIAC) announced a strategic partnership with Jus Mundi and its AI platform, Jus AI. For those practicing in the region, the announcement was more than a technological update. It reflects an institutional willingness to explore advanced research technology in support of arbitration practice in Dubai.

The development is timely. As arbitration grows in complexity and volume, expectations around efficiency continue to rise. Yet efficiency, particularly in expedited proceedings, must sit alongside due process. The availability of AI tools within the broader arbitration environment invites a practical question: how should technological assistance be approached by practitioners operating under Article 32 of the DIAC Arbitration Rules 2022?

MODERNISATION WITH PURPOSE

From a practitioner’s perspective, the attraction of platforms such as Jus AI is clear. Advanced research tools that retrieve authorities, compare jurisprudence and synthesise procedural trends in seconds are valuable in fast-moving disputes. Institutional access to such technology may assist in research and knowledge management, although procedural decisions remain governed by the DIAC Arbitration Rules and tribunal discretion.

The partnership also includes initiatives aimed at increasing accessibility to selected institutional determinations, subject to DIAC’s publication policies. Greater transparency strengthens predictability — something arbitration users value. In that sense, DIAC’s move reflects confidence in its processes.

Modernisation, however, should not be conflated with automation. Arbitration remains a human process, dependent on judgment, discretion and fairness. Technology must support those elements, not dilute them.

ARTICLE 32: SPEED REQUIRES STRUCTURE

Article 32 provides for expedited proceedings where the amount in dispute does not exceed AED1,000,000, where parties agree, or where urgency justifies acceleration. Typically conducted before a sole arbitrator, the process aims to produce a final award within three months of transmission of the file.

Anyone who has acted in expedited arbitration recognises the pressure such timelines create. Submissions are tighter. Evidence must be marshalled swiftly. Hearings, if held, are concise. The tribunal must deliberate efficiently without sacrificing analytical rigour.

Expedition does not reduce procedural obligations. Equality of arms, the opportunity to present one’s case and reasoned decision-making remain non-negotiable. An expedited award that compromises these fundamentals risks challenge at the enforcement stage.

Although Article 32 does not address artificial intelligence, the realities of compressed procedure make questions of research and preparation particularly significant.

AI IN PRACTICE: A TOOL, NOT A CRUTCH

Used properly, AI can assist practitioners significantly. It may identify overlooked authorities, flag comparative approaches and accelerate document review. In expedited proceedings, that efficiency may be decisive.

However, AI systems operate through predictive modelling and pattern recognition; they do not exercise legal judgment. Generative tools have, in some instances globally, produced inaccurate citations or over-simplified principles. In conventional arbitration, such errors might be corrected with time. Under Article 32, there may be far less opportunity to do so.

Responsibility for submissions remains with counsel, and responsibility for reasoning rests with the tribunal. AI may inform the process, but it cannot replace professional scrutiny required to test authorities, verify sources and weigh evidentiary value.

Disciplined verification is therefore essential. Authorities generated by AI must be checked. Summaries should be reviewed against primary sources. Analytical shortcuts are risky, particularly where awards must withstand external scrutiny.

EVIDENTIARY INTEGRITY AND TRANSPARENCY

Evidentiary integrity extends beyond admissibility. It concerns how evidence and legal argument are constructed. If AI tools materially shape research or drafting, the question arises whether, and to what extent, that reliance should be transparent.

Arbitration rules do not uniformly require disclosure of AI use. Nevertheless, professional guidance is increasingly attentive to the issue. In expedited proceedings — where time for rebuttal may be limited — transparency can reduce the risk of later procedural complaint.

The issue is not impropriety; AI use is becoming commonplace. The issue is confidence. Parties must feel assured that submissions are grounded in verified sources and that no technological opacity undermines equality of arms.

A prudent approach combines rigorous internal verification with measured transparency where reliance on AI has been significant.

ENFORCEABILITY AND PROCEDURAL RISK

Arbitration’s credibility ultimately depends on enforceability. Awards rendered under expedited procedures must comply with the New York Convention and applicable domestic arbitration laws. Allegations of procedural unfairness remain a common basis for challenge.

If technological tools were relied upon without appropriate verification, arguments concerning accuracy or fairness could be raised at the enforcement stage. Even if unsuccessful, such arguments may create delay and uncertainty.

Conversely, responsible use of AI may strengthen awards. Comprehensive research tools can enhance analytical depth and reduce inadvertent omission of relevant authority. The difference lies in governance. Training, ethical awareness and clear internal protocols are essential. Institutional engagement with AI tools underscores the importance of pairing innovation with discipline.

A PROFESSION IN TRANSITION

For practitioners, digital literacy is no longer optional. Understanding how AI tools function — and where they fall short — forms part of competent representation. For institutions, the challenge is maintaining legitimacy while embracing efficiency.

Article 32 reflects DIAC’s commitment to proportionality and responsiveness. Used responsibly, technological research tools may support effective preparation within its timeframes. Arbitration has always evolved. Virtual hearings, electronic bundles and online filings were once novelties; they are now standard. Artificial intelligence may follow a similar trajectory.

In expedited proceedings, the balance is delicate. Speed must not erode fairness. Efficiency must not obscure evidentiary integrity.

If that balance is maintained, AI will not threaten Article 32 proceedings. Properly governed, it may enhance them — ensuring that arbitration in Dubai remains commercially responsive and procedurally sound in an increasingly digital legal landscape.

Text by:

Abubaker Karmustaji, co-founder & head of dispute resolution practice, SAT & CO Advocates and Legal Consultants