Search

We report on a project exploring the development of an AI-enabled system for researching and verifying news articles. In particular, we underscore the value of journalists in the role of designers in a wider multi-disciplinary team including AI experts and interaction designers. We unpack our learnings by presenting three sensitizing concepts for Human-Centred AI technologies in the context of journalism. We contribute these concepts to provoke discussion and inspiration for design work.

The use of AI technology can help to automate news verification workflows, while significantly innovating journalism practices. However, most existing systems are designed in isolation without interactive collaboration with journalists. ‘DMINR’ project aims to bring humans-at-the-center of AI loop for developing a powerful tool that is sympathetic to the way journalists work. In this paper, we attempt to understand how AI can shape journalists’ practices and, crucially, be shaped by them; we aim to design human-centred AI tool that works in synergy with journalists’ practices and strike a useful balance between human and machine intelligence. In this paper, we conducted a Co-design workshop to inform the design of the ‘DMINR’ system. Based on the findings, we outline the main challenges for designing AI systems in the context of journalism, that can serve as a resource for Human-AI interaction design.