Metacognitive Interventions for Navigating the Drawbacks and Threats of Using GenAI (MIND-TUG)
Background
Generative AI (GenAI), in the form of chatbots and software support, has become firmly established in society and is now widely used by the general public. While GenAI can have desired effects for users (e.g., productivity gains), emerging research also highlights several psychological drawbacks and threats of GenAI use, including overreliance and skill decay. Yet, it remains unclear how Human–AI Interaction (HAII) should be designed to minimize such undesired effects. A promising individual-level approach to this challenge lies in a closer examination of human metacognition. Human metacognition encompasses all cognition about one’s own cognitive processes, a construct sometimes referred to as ‘thinking about thinking’. More recently, interest in metacognition has also increased within the field of HAII, as metacognitive abilities are considered essential for effective self-regulation in the context of GenAI use.
Motivation
Metacognitive interventions tailored to GenAI could, therefore, equip users with strategies to protect themselves against the psychological drawbacks and threats associated with GenAI use. However, little is known about whether everyday users can easily learn these interventions or whether they lead to measurable reductions in the undesired effects over time. It remains unclear whether metacognitive interventions shape GenAI use in situ, mitigate risks such as overreliance, or how they should be designed to reliably prevent problematic AI use across diverse user groups.
Objectives
Our MIND-TUG project aims to address this gap. A central objective is to identify interventions that are simple and intuitive enough to be applied independently by GenAI users in the general public. We will analyze promising metacognitive interventions that have already been applied in other contexts, investigate their effectiveness in the context of GenAI use through experimental testing, and adapt them to GenAI use where necessary. The project findings will then be communicated to the general public. Through this work, we aim to support the general public in improving their use of GenAI.
