The Engineer's Dilemma: A Review of Establishing a Legal Framework for Integrating Machine Learning in Construction by Navigating Precedents and Industry Expectations

TL;DR

This paper reviews how legal principles and precedents can guide the integration of machine learning into engineering, emphasizing the importance of legal-technical interplay for responsible adoption.

Contribution

It proposes a legal framework based on analogical reasoning and precedents to help engineers navigate liability and accountability in ML deployment.

Findings

Legal precedents support ML integration in engineering.

Analogical reasoning aids in embedding ML within existing codes.

Guidelines from legislation and standards can facilitate responsible ML use.

Abstract

Despite the widespread interest in machine learning (ML), the engineering industry has not yet fully adopted ML-based methods, which has left engineers and stakeholders uncertain about the legal and regulatory frameworks that govern their decisions. This gap remains unaddressed as an engineer's decision-making process, typically governed by professional ethics and practical guidelines, now intersects with complex algorithmic outputs. To bridge this gap, this paper explores how engineers can navigate legal principles and legislative justifications that support and/or contest the deployment of ML technologies. Drawing on recent precedents and experiences gained from other fields, this paper argues that analogical reasoning can provide a basis for embedding ML within existing engineering codes while maintaining professional accountability and meeting safety requirements. In exploring these issues, the discussion focuses on established liability doctrines, such as negligence and product liability, and highlights how courts have evaluated the use of predictive models. We further analyze how legislative bodies and standard-setting organizations can furnish explicit guidance equivalent to prior endorsements of emergent technologies. This exploration stresses the vitality of understanding the interplay between technical justifications and legal precedents for shaping an informed stance on ML's legitimacy in engineering practice. Finally, our analysis catalyzes a legal framework for integrating ML through which stakeholders can critically assess the responsibilities, liabilities, and benefits inherent in ML-driven engineering solutions.