Reassessing Intellectual Property Valuation: Economic Models for AI-Generated Assets and Patent Portfolios

Abstract

The rapid evolution of generative artificial intelligence (AI) has reshaped the landscape of intellectual property (IP) by introducing a new class of intangible assets—AI-generated creations. These assets challenge traditional economic valuation models that were designed for human-generated works and discrete patents. Conventional approaches such as discounted cash flow, market comparables, and option-based frameworks often fail to capture the unique characteristics of AI-generated assets, including ambiguous authorship, rapid obsolescence, near-zero marginal reproduction costs, and high interconnectivity within patent portfolios. This paper reassesses intellectual property valuation by proposing refined economic models tailored for AI-driven innovation. It examines the limitations of traditional valuation frameworks and introduces integrated models that incorporate stochastic revenue forecasts, legal attribution risks, real-options flexibility, and portfolio interdependencies. Furthermore, it explores how data-driven techniques and machine learning can enhance valuation accuracy and transparency. The study also evaluates the implications for accounting, taxation, mergers and acquisitions, and public policy, suggesting that a unified framework linking law, economics, and computation is essential for valuing AI-generated assets and patent portfolios in the digital era.