CEC04-05 Feasability of quality criteria and tentative assessment tools to align with artificial intelligence

Abstract

Artificial Intelligence (AI) tools are the emerging technology that is on everyone's lips nowadays – AI chat bots manage customer support, ChatGPT summarises long texts, Copilot assists with programming … But are these tools also capable enough to be of use in science in general and regulatory toxicology in particular? Could AI also help with the critical appraisal of toxicokinetic studies?

The session will provide an overview on AI in the broader sense, on possible applications, and related challenges. It will also critically address concerns that arise from the use of this technology in scientific settings, such as ethical considerations, financial and ecologic costs, transparency and data protection as well as inherent biases.

The participants of the CEC will have the opportunity to test the (non-AI) tool developed by our group to assess the reliability of kinetic studies on two example publications during the practical exercise. After this session's talk, we will take an interactive look at one available AI tool to see how the tool performs with this same task on one of the example publications. Thus, participants will be able to directly compare their results with results generated by the AI tool and to gain first-hand experience of the advantages and disadvantages of using AI in the systematic evaluation of kinetic studies.