Human-Machine Collaboration for Strategy Foresight: The Case of Generative AI

• 成果类型: Article; Early Access
• 署名作者: Picavet, Marc E. B.; Maroni, Peter; Sandhu, Amardeep; Desouza, Kevin C.
• 署名单位: Queensland University of Technology (QUT)
• 刊物名称: PUBLIC ADMINISTRATION REVIEW
• ISSN/ISSBN: 0033-3352
• DOI: 10.1111/puar.70048
• 发表日期: 2025

摘要

Generating strategic foresight for public organizations is a resource-intensive and non-trivial effort. Strategic foresight is especially important for governments, which are increasingly confronted by complex and unpredictable challenges and wicked problems. With advances in machine learning, information systems can be integrated more creatively into the strategic foresight process. We report on an innovative pilot project conducted by an Australian state government that leveraged generative artificial intelligence (AI), specifically large language models, for strategic foresight using a design science approach. The project demonstrated AI's potential to enhance scenario generation for strategic foresight, improve data processing efficiency, and support human decision-making. However, the study also found that it is essential to balance AI automation with human expertise for validation and oversight. These findings highlight the importance of iterative design to develop robust AI tools for strategic foresight which, alongside stakeholder engagement and process transparency, build trust and ensure practical relevance.