Machine Learning Model Assets

Definition

Trained machine learning models — including their weights, architectures, and associated deployment infrastructure — recognised as proprietary intangible assets with commercial value. The value of a machine learning model asset depends on its task performance, uniqueness, the cost and scarcity of replication, the defensibility of its training data, and its integration into revenue-generating products or processes. Model assets are increasingly material in technology M&A transactions.