How Is AI Changing The World of Arbitration? – Artificial Lawyer

Artificial Lawyer asked Roberta Downey, Partner and Head of Vinson & Elkins’ International Construction group, what she thinks about how AI will change the world of arbitration and the ethical aspects that need to be considered. Here’s what she said.

How are arbitrators currently integrating AI tools into their practices, and what is the state of adoption within the field today?

I do not sit as an arbitrator personally, but would be surprised if many arbitrators themselves are integrating it into their practices beyond remote hearings and navigating databases of electronic evidence and legal submissions. 

In my capacity as acting as counsel in arbitrations, however, the pandemic certainly accelerated the use of technology in resolving disputes. The areas where AI is used and, I believe, there is more potential are those aspects of the work where there is repetition, or managing large volumes of evidence, or translations for international projects. For example, we use AI:

to help search through litigation databases – there is even a legal precedent endorsing the use of technology assisted review (“TAR”) – see Pyrrho Investments Ltd v MWB Property Ltd [2016] EWHC 256,

to put together trial bundles (which are almost always electronic or have an electronic element now), and

to translate documents from multi-national projects (eg we have a project with a Chinese client performing work on a European project with native Greek speakers for an arbitration that will be conducted in English in London). team some arbitrators require the submissions and evidence that is filed to be hyperlinked to underlying records. 

I have also used artificial intelligence to give arbitrators a “virtual” site visit – creating digital models of the construction site and showing how the work progressed over time.

But I am sceptical about whether good lawyers will use AI to draft submissions, at least in the kinds of disputes we manage at Vinson & Elkins. 

This is because the issues that arise are complex and will depend on the unique facts and circumstances of each case, and construction law is spread across so many sources and areas. E.g. only last year the English Supreme Court in a tax case issued the leading decision on the scope of a developer’s right to change the work under a standard form (JCT) contract: see the decision in R(Cobalt Data Centre 2 LLP v Revenue and Customs Commissioners [2024] 1 WLR 5213.  It is hard to see how AI might pick up that sort of authority.  Indeed, there have been some highly publicised and embarrassing mistakes when lawyers have relied on AI to draft submissions and the AI tools have hallucinated.

The way disputes are resolved is also a factor because it limits the publicly available knowledge / learning that is needed to train AI – in the construction industry, so many disputes are resolved in confidential processes like adjudication, arbitration, mediation. 

In contrast, the use of AI in the construction industry is already widespread and increasing:  in addition to document management systems (most projects have shared databases nowadays and we invariably use a litigation database on big disputes), it is filtering through all stages of a construction project – be it tendering, risk spotting, contract management, managing procurement / the supply chain, planning /programming the works to maximise efficiency, capturing the records, preparing as-built drawings and models etc.  Some of the tools are:

BIM (Building Information Model) and digital twins to create digital representations of physical assets, functional processes and systems for a building or infrastructure report;

Automated construction such as 3D printed buildings;

So called “smart contracts” which can help with management of supply chains, streamline payment process.

To what extent is it feasible for an AI system, with current technology, to arbitrate a case to a successful conclusion?

I have read about Online Dispute Resolution (ODR) using technology to resolve disputes rather than go to Court but have no experience of it, know nobody who has used it and am sceptical about its suitability for the kind of work that my peers and I do.  I can see the attraction for simple, low value disputes but the nature of the work I do is complex, high value and high stakes – I cannot see parties on either side of the dispute being comfortable to delegate the decision making in such a dispute to a machine.

I note that the EU has decided to discontinue its ODR Platform from 20 July. I understand that was driven by low use and even lower acceptance of the system

What ethical considerations arise from allowing AI to serve as an arbiter in legal disputes?

In some jurisdictions / seats, the procedural laws make it mandatory for the arbitrator to be human – eg the French Code of Civil Procedure (Art 1450), the UAE’s Federal Law No 6 of 3 May 2018 (Article 10(1)), Turkey’s International Arbitration Law No 6686 of 21 June 2001 (Article 7(B)(1), the Dutch Code of Civile Procedure (Art 1023), the Scottish Arbitration Rules / Arbitration (Scotland) Act 2010 (Sch 1, rule 3) and Brazil’s Law No 9.307/96 (Article 13).

The English Arbitration Act does not expressly prohibit it by stating the arbitrator must be human but there are other provisions which would be incompatible with having an AI arbitrator – eg the authority of an arbitrator being personal and ceasing upon death (s. 26(1)) – and others where it is hard to  see how an AI arbitrator could deal with the more subtle powers and duties of an arbitrator where there is more scope for nuance and subjectivity such the duty to impose duties of fairness and impartiality, the exercise of discretion

Could AI contribute to speeding up the arbitration process, and thereby reduce costs?

Yes – as stated above, we have seen this with the pandemic – more remote hearings make it easier to convene tribunals and parties.  We have also seen reduced trial lengths – whereas 4-6 week trials were common place on major construction disputes, it is now rare for tribunals to agree to trials that last more than 2 weeks.  That is against a background where the volume of documentary evidence has increased exponentially and the complexity of the issues that arise in the disputes have not got any simpler.

—

Thanks to Roberta Downey, who is Head of Vinson & Elkins’ International Construction group. She is based in the London office with close to 30 years of experience in commercial dispute resolution, including large trials and highly technical disputes.

You can see her full bio here.

—

Legal Innovators Conferences in New York and London – Both In November ’25

If you’d like to stay ahead of the legal AI curve….then come along to Legal Innovators New York, Nov 19 + 20, where the brightest minds will be sharing their insights on where we are now and where we are heading. Also – worth noting that the London conference will have a special day devoted to litigation on November 4th – see link below.

And also, Legal Innovators UK – Nov 4 + 5 + 6

Both events, as always, are organised by the awesome Cosmonauts team! 

Please get in contact with them if you’d like to take part.