This essay is based on a survey study of teachers in upper secondary school in Skåne,Sweden. The questions the teachers have answered have pertained to how they assess theirstudents with regards to the abilities they are supposed to develop, how these assessmentscorrelate with the relevant documents from Skolverket, and how the continuous growth of AI,and students getting access to these AI:programs, have impacted how teachers assess theirstudents. Our findings indicate that teachers rely heavily on written exams, and do not allowtheir students the varied forms of assessment that Skolverket demands that teachers shouldoffer the students. We discuss possible effects of this, how this relates to theories of reliabilityand validity, and what the future might look like for teachers moving forward with the risingusage of AI in school.