Artificial intelligence (AI) technologies have long been used in the creation and distribution of journalism, but their application has largely been limited to well-resourced, large news organizations. With the emergence and popularization of generative AI systems, such as ChatGPT, the barrier of use of AI technology has lowered and this dissertation considers how the technology is reshaping, and has the potential to further transform how journalism is produced and consumed within small news organizations. Through a design science informed approach, this topic is explored through the development of a generative AI chatbot prototype, which served as a research instrument during interviews with small newsrooms staff from Canada and the United States. These interviews informed the redevelopment of both the open-source version of the prototype and the design of simple implementation documentation. This study analyzed what use of the prototype revealed about small newsroom reception to generative AI technologies, and highlighted what considerations shape and moderate its use. This project explored how generative AI technologies might alter small newsroom-audience relationships and provides insight into potential user reception to the technology. Finally, this research works to determine how the competitive landscape and overall future of small newsrooms might be impacted by generative AI technology. Through a review of prevailing journalism, artificial intelligence, and innovation research, this dissertation examines how generative AI systems, including chatbots, are one in a long line of innovations that have transformed journalism practice, contextualizing its potential impact on small newsrooms. The findings of this work indicate that small news organizations are open to generative AI technologies, particularly as an administrative support tool, as they offer efficiencies and opportunities to engage with audiences that would not otherwise be possible given resource constraints unique to small news organizations. These constraints make generative AI implementation a riskier endeavor for small news outlets compared to larger, more resourced news organizations. This use is guided by adherence to journalistic norms, particularly transparency and accuracy, to protect the distinctive relationships small news organizations have with their audiences.
GitHub Repository for the Chatbot Prototype