This chapter takes stock of the questions of trust and responsibility for the use of AI/ChatGPT that have followed the book’s examples and analyses. The question of responsibility, which initially revolves around ChatGPT/Chat’s authorless and sourceless text universe, is posed openly and exploratively, but also concretely confrontingly. And the ways in which Chat asserts its own abilities, attributes/disclaims responsibility for its own proposals, embeds authoritative voices, makes normative turns and, most recently, genre plagiarisms, call not only for the question of responsibility, but for questions of trust and credibility. In 2025, the combined questions of responsibility and trust will of course not only be asked of Chat/ChatGPT, but of the entire range of GAI technologies. And it is to be expected that researchers, teachers and students, as academic users, will have an ‘own responsibility’. But the consequence of the shift in responsibility is surprising. The GAI tool and ChatGPT as a text producer will be held free from responsibility. In addition, the issue of trust and credibility is slipping out of sight. And the party that must take a position now seems to have shifted to the students. In a set of regulations with resonances both in the EU/DK and in the universities’ proposals, the responsibility is made clear to the individual student: ‘It is your own choice whether you want to use GAI in projects and theses, but when you use GAI, you must be critical…’ The universities also make an appeal, namely that ‘we must move away from cheating and towards good academic practice’. But even though this ‘we’ applies to all of us, the responsibility for interpreting what is up and down about AI, cheating and good academic practice was shifted to the students! The shifts in responsibility are the debate and analysis’s final, but also currently critical, focal point. Here ‘in the midst of the critical’, I am also rounding off the book with a status and position, spring 2025.