Typical weaknesses in AI generated texts

Detection tools and large language models are undergoing such rapid and mutually reinforcing development processes that the use of generative AI in take-home assignments cannot usually be proven. More important than the question of whether generative AI has been used – even when prohibited – in written assignments is often the question of whether the resulting texts meet academic standards. The following catalogue of criteria can be helpful in assessing texts and talking to students.

Downloadable files

Typical_weaknesses_in_AI_texts.pdf

Related materials

Typische Schwächen KI-generierter Texte