Are Societal Promises in Science and Technology Substantiated? A Study of Value Expressions in Patents

Abstract

This study examines whether societal promises in patents are substantiated by the technological capabilities of the inventions they describe. We analyze value expressions (VEs)—narrative statements about societal or commercial benefits—in 175,730 AI and nanotechnology patents (2005–2023). Combining GPT-4o and RoBERTa Large models, we categorize these as public value expressions (PVEs), private value expressions (PRIVEs), or combined expressions (PVE-PRIVEs). We then test whether patents containing these narratives have technical features aligned with addressing societal challenges, as measured by mapping patents to UN Sustainable Development Goals (SDGs) based on their technical classifications. Our findings reveal that patents containing PVEs are significantly more likely to have technological capabilities that address societal challenges, while those with PRIVEs show the opposite pattern. Notably, patents articulating both public and private values (PVE-PRIVEs) show the strongest alignment with SDG-mapped technological capabilities. These results suggest that value-based statements in patents are not merely rhetorical but meaningfully reflect or signal an invention’s potential impact. This research contributes to ongoing debates about the credibility of societal promises conveyed through value-laden language in science and technology.