Large Language Models have been integrated in the prominent analysis programmes MAXQDA or ATLAS.ti. Other programmes such as canvs.ai, AILZYE, Cody or CoLoop are on the way. This opens up many fields of application for the analysis of texts with content analysis methods: for example, summarising coded segments or suggesting subcategories.

But can such Large Language Models be used in the analysis steps of content analysis procedures at all? Does the logic of the algorithms fit the logic of human analysis steps? Is it a complement or does it limit human, creative work? Is the use of Large Language Models justifiable with methodological foundations that are based, among other things, on communication theories that address communication between humans and not between humans and machines?

The discussion is open! I am looking forward to your contributions.