| dc.description.abstract | Introduction: Artificial intelligence innovation, particularly by start- and scale-ups promises solutions to grand societal challenges and substantial financial returns. Despite substantially lowered barriers for entry to artificial intelligence development, uncertainties about intellectual property have inhibited innovative activity from accelerating. Past research has shown how well-designed intellectual property strategies tackle uncertainties, in turn amplifying innovation. Well-researched in related fields; software start- and scale-ups, little is known on intellectual property strategies utilised in the context of artificial intelligence. This abductive study is the first to investigate intellectual property strategies and the effects of innovation and market factors on its design in the context of artificial intelligence start- and scale-ups.
Theory: Extent research on software start- and scale-ups provided seven appropriability mechanisms: copyrights, database rights, trade secrecy, secrecy, lead-times, complementary assets, and technological complexity. Intellectual property strategies are affected by six factors, which differ between software and artificial intelligence. The type of key resources, vital to product development. Tacitness being the degree to which knowledge relies on skill and expertise Ambiguity being the observability of cause-and-effect relationships in knowledge. Open-source asset is the extent to which products rely on open-source assets. Market newness is the novelty of consumer needs and distribution channels to firms. R&D intensities is the size of research and development investments in relation to total investments.
Methods: The ways in which factors influence the usage of appropriability mechanisms utilised by artificial intelligence start- and scale-ups was investigated through cross-sectional qualitative data from nine semi-structured interviews, seven with chief executive officers, two with specialised artificial intelligence intellectual property advisors. Thematic analysis was employed to explore causal relationships between factors and appropriability mechanism usage.
Results: Findings confirmed industry similarities to validate the abductive approach and corroborated five theory-driven themes. R&D intensities provided contradictory results, confounded product maturity.
Discussion/Conclusion: The propensity for secrecy is concluded largest, albeit hampered by open-source motivations. Lead-times become increasingly relevant to keep up with rapidly changing consumer needs. As data and related assets increase in relevance, so does interest in database rights. However, artificial intelligence start- and scale-ups are deterred from its usage due to unclear legislative definitions. Insights provide a springboard for further research on intellectual property strategies in the context of artificial intelligence. Policy makers can improve database rights based on findings, better defining effort and the eligible of synthetic or derivate datasets. Uncertainties are addressed by examining their sources and illustrating the suitable strategies. | |
