Dilemma of Human-AI Collaborative Innovation: Exploring the Boundaries of Human-AI Reciprocity and Platform Incentive Mechanisms in Open Innovation

Artificial intelligence (AI) aids many open innovation platforms in human ideation and evaluation. While current research highlights AI's innovation potential, it has paid less attention to the ethical challenges from information flow in human-AI collaboration. This flow causes a tradeoff between generating novel ideas and ensuring their accurate evaluation free from strategic manipulation. Using an evolutionary game model for open innovation’s design and evaluation stages, this study explores HAI reciprocity’s impact on interpersonal dynamics and the interplay between platform incentives and AI. Results show constraints are crucial to sustaining positive cycles of HAI reciprocity. Unconditionally cooperative AI slightly boosts group innovation but reduces individual human performance and hinders valid evaluation. In contrast, Reputation-based AI improves both human and population performance. With fixed incentives, having more AI increases system stability and improves innovation outcomes. Platforms should integrate AI into mechanism design, rather than treating it as a technical tool.