New EU AI Act set to have global implications

The EU AI Act will standardise the use of AI across Europe and is likely to set a global precedent. Image: The Decoder.

New EU regulations on artificial intelligence are set to impact the global construction sector, say Sheena Sood and Sean O’Halloran of construction law firm Beale & Co.

The digital transformation of the global construction industry has significantly accelerated with the adoption of artificial intelligence (AI), enhancing operational efficiencies, safety, and resource management.

The use of AI in the European Union is poised for pivotal changes with the recent ratification of the EU AI Act. With the European Parliament’s approval and its impending formal adoption, this groundbreaking legislation, marking the first comprehensive attempt to legislate AI globally, will standardise the use of AI across Europe and is likely to set a global precedent.

The value of AI in construction has been underscored by significant advancements in automated risk assessments, safety monitoring and efficient resource allocation. This technology has notably enhanced the accuracy of project timelines, budget forecasts and compliance with stringent safety regulations. Moreover, AI-powered surveillance systems have been increasingly deployed on construction sites to bolster health and safety compliance and mitigate risks associated with theft and trespassing.

Proposed by the European Commission in April 2021, the intention behind the AI Act is to lower the risks that AI poses to the fundamental rights of citizens, whist also fostering responsible innovation within a clearly defined legal framework. The AI Act has been drafted as a regulation and as such it will create uniform AI governance across every EU member state.

The AI Act defines AI systems and categorises AI applications based on their risk levels to human safety and fundamental rights into four tiers – unacceptable risk, high risk, limited risk and minimal risk.

Systems posing unacceptable risks, such as real-time biometric identification technology in CCTV, will face outright bans. High-risk applications, including AI used in critical infrastructure and worker safety, will require registration and continual assessments. Applications with limited risk must fulfil transparency obligations to inform end users that they are interacting with AI, whereas those categorised as minimal risk will be subject to minimal regulatory constraints.

High-risk AI systems will be subject to stringent standards, including meticulous record-keeping, human oversight and performance metrics, ensuring their deployment is both transparent and accountable.

Following its expected enactment in May or June 2024, the AI Act will introduce phased compliance deadlines for various AI applications, with high-risk and general-purpose AI systems facing specific timelines for adherence. Non-compliance will attract severe penalties, potentially up to €35m or 7% of the entity’s global turnover, whichever is greater.

Echoing the ambitions of the General Data Protection Regulation (GDPR), the EU AI Act seeks to balance innovation with rigorous regulatory oversight, aiming to set a global standard for AI governance. As the European construction industry prepares to navigate this evolving technological and regulatory landscape, firms will need to evaluate and adapt their AI-driven processes to ensure compliance.

Sheena Sood is a senior partner and Sean O’Halloran a senior associate at the construction law firm Beale & Co.

Click here to read more about the EU AI Act on the Beale & Co website.