Economic Value of AI Innovation: An Empirical Exploration of U.S. Patents

Abstract

While recent research has established that AI innovations tend to be more incremental and process-oriented than comparable IT innovations, it remains unclear whether these characteristics translate into differential economic value. This emergent research extends the comparative analysis of AI and IT patents by incorporating a stock-market-based economic valuation measure. Using the USPTO AI Patent Dataset and PatentsView data, we construct matched samples of AI and IT patents and compare their estimated private economic values (in 1982 million dollars). Preliminary analysis suggests that despite being less radical, AI patents command higher market valuations than IT patents, potentially reflecting the strategic importance of AI as a source of competitive advantage. This research-in-progress contributes to the emerging discourse on AI innovation management by adding an economic value dimension to the form magnitude framework established in prior work.