Minimising disagreement is valuable from both financial and relationship perspectives. Financially, as Lord Denning observed, the occurrence of conflict is ‘One of the greatest threats to cash flow’. Litigation is ‘Frequently lengthy and expensive. Arbitration…is often as bad'. From a relationship standpoint, a practitioner notes that ‘ill-will’ often escalate before cases reach mediation. Dispute prevention therefore increases the likelihood of preserving commercial relationships by avoiding blame.
How can Artificial Intelligence (“AI”) be integrated into current dispute prevention approaches to address legal risks? This post explores the question using the construction industry as a test case. Characterised by multi-party contracting, bespoke project designs and inevitable uncertainty of on-site conditions, the risks of disputes can be quite significant in the construction industry.
This post will outline how AI tools can be applied to enhance current prevention approaches in the industry through contract drafting, design specifications and the use of Dispute Review Boards (DRB). We consider the potential to achieve ‘dispute attrition’ by integrating AI, to both reduce the risk of disagreement and to prevent disagreements from unfolding into formal disputes.
References to AI in the following points to (a) Machine learning, which trains systems to learn from data; (b) Natural language processing (NLP), which allows computer systems to process and analyse vast amounts of natural language and speech; and (c) Deep learning through artificial neural network layering, which is a subset of machine learning that reduces feature-engineering required for data preparation, with greater potential to emulate the capacity to reason through pattern and relationship recognition. These capabilities are enabled by ‘Big data’, which is distinguishable from traditional data with greater volume, velocity and variety.
Project planning – Legal risk management
Legal contracting is the cornerstone of dispute prevention, which minimises conflict by specifying the obligations and liabilities of participating stakeholders. Potential for conflict arises because contracts are inevitably incomplete due to the futurity problem of unforeseen contingencies, and high transactional costs otherwise incurred to specify ex-ante parties’ obligations in all possible outcomes, most of which will not realise.
The potential for conflict is highlighted by the Centre for Public Resources, with ‘ambiguous contract documents'and failure to ‘deal promptly with… unexpected conditions’ being common sources of conflict. We suggest that machine learning can enhance the clarity and scope of ex-post gap-filling contract mechanisms, thereby minimise differences stemming from contract interpretation, particularly with modified standard clauses.
Two examples are notable. Firstly, extension of time and variation clauses define events that merit adjustments of completion date due to delays beyond the contractor’s control. Contract analytic platforms that deploys NLP tecniques can search through precedent to ensure the clarity of language and minimise the risk of human error.
Secondly, legal contracting can be complemented with tools that allow for cost predictions, based on historical data of projects with comparable complexity, size and location. This is complimentary to Building information models (“BIMs”), which can improve design precision and minimises unforeseen obstacles.
Project execution – Dispute prevention
Following project commencement, periodic meetings and document exchange between parties are indispensable to monitor progress. However, this carries an inevitable risk of human error, including subjects that may be undetected or omitted. The integration of NLP in project management can facilitate the extraction of key data from email correspondences, construction details from the BIM, and a more comprehensive understanding of progress through data visualisation.
When disputes arise, DRBs are commonly utilised to prevent escalation. These bodies draw attention to the likely causes of conflict, such as milestone delays and facilitate a conversation on revised courses of action. The value of DRBs is evident from the fact that over 10 years, only 25 disputes with the involvement of DRBs have escalated to further dispute resolution, and 841 have been resolved.
In the foreseeable future, the use of AI tools can aid DRBs with the ability to recall information expeditiously. Big data allows for efficient access to a greater volume of data. One challenge against further integration of AI is the lack of the ‘human touch’, where decisions are limited to historical data. This concern however can be addressed where algorithms are only given responsibility, not authority, to provide observations amenable for parties’ consideration. It therefore has the potential to foster a non-adversarial and collaborative environment whilst removing concerns of human error.
Dispute prevention confers substantial commercial and relationship benefits conducive to parties engaging in multi-party contracting scenarios. The integration of AI can encourage both minimisations of risks and early anticipation of disagreement, which offers a useful toolkit for conflict avoidance.