HFW instructed in landmark Singapore AI dispute involving training data and proprietary design content

Global law firm HFW has been instructed in what is believed to be the first major artificial intelligence (A.I.) dispute of its kind to be brought in Singapore, acting for a Singapore-headquartered A.I. company as cross-border dispute resolution strategist in a high-stakes dispute against an Asian company in the engineering and diversified industrials sector. The matter is led by Partner Shaun Leong and assisted by Senior Associate Theodore Ang.

The dispute centres on allegations that proprietary data was used without authorisation to train and develop advanced A.I. models, raising novel and complex questions at the intersection technology and liability in the age of artificial intelligence.

At the heart of the case is an A.I. powered system designed to automate the integration and clash resolution of complex design drawings, including architectural, structural and mechanical & electrical engineering plans across large-scale projects. The system enables multidisciplinary project teams to identify and resolve clashes digitally, delivering clash-free, optimised design outputs at speeds reported to be up to eleven times faster than traditional manual coordination methods. This technology represents a step-change in how major construction and infrastructure projects are designed, coordinated and delivered.

The dispute mirrors a growing wave of high-profile A.I. training-data litigation unfolding globally. These include widely reported claims against OpenAI, where publishers, authors and content owners have challenged the use of proprietary materials to train large language models such as ChatGPT, and the recent Disney and Universal v MiniMax proceedings, which focus on whether creative works may be ingested to train generative A.I. systems without consent, even where the outputs are transformed or purposively distinct.

Like those cases, this Singapore dispute touches on a fundamental and unresolved question in A.I. law:

Can A.I. developers lawfully use proprietary materials to train and develop their models without permission, particularly where the A.I. output differs from the original content and generates new meaning, function or separate commercial value?

The case therefore sits squarely within the global debate on training-data rights, transformation, derivative use, and the allocation of risk between A.I. developers, technology users and content owners. Its outcome has the potential to influence not only future disputes in Singapore, but also how A.I. systems are developed and deployed across Asia and beyond.

Shaun Leong, Partner, HFW:

“We’re proud to be advising on this cutting-edge, precedent-setting dispute, which underscores the firm’s position at the forefront of A.I. liability, technology disputes and cross-border dispute resolution strategy.

As A.I. adoption accelerates across safety-critical and asset-heavy sectors, disputes of this nature are expected to become increasingly central to the global disputes landscape – and it’s a privilege to help shape that future.”