This paper asks whether AI makes impact assessment more sustainable or simply faster, and examines its implications for judgment, participation, trust, and decision quality.
Artificial intelligence is increasingly being introduced into impact assessment to accelerate evidence synthesis, automate reporting, support prediction and improve monitoring. Yet greater efficiency does not necessarily lead to better or more sustainable decisions. This paper critically examines whether the growing use of AI in impact assessment improves decision quality, transparency, participation, and long-term sustainability, or primarily makes existing processes faster. It distinguishes between efficiency, effectiveness and sustainability as three different tests of AI-enabled impact assessment. The paper considers where AI can strengthen practice through improved access to evidence, scenario exploration and adaptive monitoring, while also examining risks related to bias, opacity, weak contextual understanding and over-reliance on automated outputs. Particular attention is given to professional judgment, stakeholder participation and accountability, areas where technological efficiency may conflict with the broader purposes of impact assessment. The paper argues that the value of AI should be judged not by how much assessment work it can automate, but by whether it improves the quality, legitimacy and sustainability of decisions. It concludes by proposing key questions that practitioners and institutions should ask before adopting AI across the impact assessment process.
Dr. Sarfraz Aslam is a Professor and Associate Dean (Research) at UNITAR International University, Malaysia, with expertise in responsible AI, sustainability, and education.