Associations Between Generative AI Use and Facial Expression-Derived Central Executive Network Indices: A Pilot Study

Area of Science:

• Cognitive Neuroscience
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Generative AI's rapid growth raises questions about its cognitive effects, especially in creative processes.
• Facial expression indices previously linked to brain networks (Default Mode Network/DMN, Central Executive Network/CEN) were used to explore AI's influence.

Purpose of the Study:

• To investigate task-specific associations between generative AI use and facial expression-derived neural network indices.
• To explore potential correlations between creative self-efficacy and these neural network responses during AI-assisted tasks.

Main Methods:

• Thirty-three business professionals participated in three AI-supported writing tasks: concept generation, concept combination, and a mixed task.
• Facial expression data was collected and analyzed to derive indices correlated with DMN and CEN activity.

Main Results:

• A significant reduction in the CEN-related facial expression index was observed during the concept combination task.
• No significant changes were found for the concept generation or mixed tasks.
• Increased creative self-efficacy correlated with less pronounced reductions in the CEN-related index.

Conclusions:

• Findings suggest generative AI may influence the CEN during specific creative sub-processes like concept combination.
• Results are preliminary due to indirect measures and lack of a control group; further research is warranted.
• Future studies should employ controlled designs and direct neuroimaging to confirm these hypothesis-generating findings.