Paper short abstract:

This presentation will address trust and trustworthiness of scholarly information in light of the proliferating uses of generative AI at every stage of knowledge production and the development of open science.

Paper long abstract:

Scholarly publications have been regarded as certified and trustworthy information. It is generally agreed that researchers abide by the Mertonian norms that prioritise public interests rather than self-interests. The calls for open access and open research adhere to the techno-utopian vision that scholarly information should be publicly accessible in a democratic society. In truth, scholarly publishing diverts from these norms and ideals for a multitude of reasons. The highly competitive academic environment, the stronghold of commercial publishers and data providers, the audit culture, and the influences commercial and political interests can tamper with scholarly pursuits and knowledge production. There exist hijacked and predatory journals, there are also increasing incidents of retraction and reports of fabricated and falsified data, stolen images, and paper mills. Generative AI, alongside the drive for open science, will accelerate and complicate scholarly publishing for there is no existing norms or regulations that forbid the uses of AI-generated contents including data, images, and texts, nor are there rules to prevent existing scholarly publications to be raw materials for large language models (LLMs). Scholarly publications are not immune from the contamination of misinformation and disinformation–and their status as certified and trustworthy information can be put into doubt with the consequence of losing credibility and public trust. In this presentation, I will address trust and trustworthiness of scholarly information in light of the proliferating uses of generative AI at every stage of knowledge production and the development of open science.