Measuring the ROI of AI is challenging due to the complexity of business processes, long timelines for realizing benefits (such as drug development), and difficulties in establishing clear productivity baselines. Companies often face hidden and indirect costs related to AI deployment and find that efficiency gains may not translate directly to cost savings or revenue increases because work expands to fill available time and some industries’ business models (e.g., billable hours for lawyers) complicate quantification. Tools like AI-powered process mining and digital twins help reveal process inefficiencies and enable more precise tracking, but comprehensive ROI assessments remain elusive and may take years to materialize.
https://www.cio.com/article/4183502/why-is-it-so-hard-to-measure-the-roi-of-ai.html