Of course digital technology such as BIM and concepts such as "digital twins" are now mainstream in construction projects, making the transition to AI-powered systems an unsurprising and generally welcomed evolution in the realm of building modelling. There is much optimism on the benefits of AI for contract management, particularly for smaller businesses with less manpower.

A wider use of AI-tools in the industry gives rise to new challenges too. The Artificial Intelligence Report published by the Royal Institute of British Architects ("RIBA") highlights the copyright of architectural designs as a key concern. Many of these issues stem from a lack of governance. On a macro level, legislation and regulation have progressed slower than the pace of technological advancement in this area. At the time of writing the UK has just signed the first international treaty governing the safe use of artificial intelligence: The Council of Europe's Framework Convention on Artificial Intelligence and Human Rights, Democracy, and the Rule of Law (CETS No. 225). The UK government has separately launched the AI Opportunities Action Plan, and announced plans to introduce legislation targeting AI models in the King's Speech. 

However, governance also extends to appropriate oversight and monitoring of AI-based tools within a business or an industry. Accompanying the astonishing and rapid ascent of AI technology has been a growing awareness that human oversight is necessary to ensure accuracy, quality, and ethical use of data. There is a general consensus that AI is only as good as i) the information fed into it; and ii) the human element which oversees it. The adage "rubbish in = rubbish out" stands firm.

In the context of formal dispute resolution, the increased use of digital technology in general has also lead to sophisticated attempts of fraud. This was brought quite spectacularly to the fore in the litigation sphere with the case of Contax Partners Inc BVI v Kuwait Finance House (KFH-Kuwait)[1] earlier this year. Contax BVI, an oil and gas company, sought leave from the Commercial Court under s.66 of the Arbitration Act 1996 to enforce a £70 million Kuwaiti arbitration award against banking group Kuwait Finance House ("KFH"). The Court granted leave to enforce the operative part of the award and entered judgment accordingly, following which Contax obtained interim third party debt orders which froze KFH's various bank accounts. It was subsequently discovered that the Kuwaiti arbitral award was in fact a fake. Not only was the award fraudulent, but supporting documents including the arbitration agreement, witness statements, and a prior Kuwaiti court decision had also been forged. In fact, no arbitration had ever taken place.

AI-altered images, videos, and documents present new risks to the reliability and veracity of evidence. A number of recent high profile cases have highlighted the dangers of using AI language processing tools such as Chat GPT to draft pleadings. Lawyers and litigants in person have been caught out by a failure to spot non-existent cases cited by ChatGPT as authority in court actions in numerous jurisdictions, including the U.K and the U.S. Dubiety of the authenticity of digital evidence has been employed by defence lawyers in the US Capitol insurrection trial of United States v Reffitt[2] in what has been coined the "deepfake defence".

Although not construction disputes, these cases highlight the gap between the current systems and human skill set and the capability of AI at the hands of actors who seek to use it for underhand or nefarious purposes.

The high volume of documents in a typical construction project mean that while AI is a welcome development in areas such as contract management, the sector is vulnerable to fraudulent activity or misuse of information. Ensuring that those responsible for the use and oversight of AI-based tools, including the courts and dispute tribunals, have the necessary skills is going to be key to its successful use in both construction projects and construction disputes.

[1] [2024] EWHC 436 (Comm

[2] 1:21-er-32 (D.D.C)