AI tools have led to a mismatch between current teaching and learning practices and their evaluation. Arguing that this enables the performance of learning without actual knowledge, Thorsten Fröhlich suggests higher education should prioritise the ‘meta-cognitive’ skills needed to understand and evaluate the outputs of generative AI.

Are we teaching students AI competence or dependence?


In October 2025, the London School of Economics Impact Blog published my latest article:

Are we teaching students AI competence or dependence?

AI tools have led to a mismatch between current teaching and learning practices and their evaluation. Arguing that this enables the performance of learning without actual knowledge,  Thorsten Fröhlich suggests higher education should prioritise the ‘meta-cognitive’ skills needed to understand and evaluate the outputs of generative AI.   

The piece draws from my competencies framework,

A Mission Statement for Higher Education in the Age of Generative AI

AI tools have led to a mismatch between current teaching and learning practices and their evaluation. Arguing that this enables the performance of learning without actual knowledge, Thorsten Fröhlich suggests higher education should prioritise the ‘meta-cognitive’ skills needed to understand and evaluate the outputs of generative AI.


🔗 Blog post: Are we teaching students AI competence or dependence?

Feel free to share it with anyone involved in supervision, academic skills training, or research education. Discussion welcome.