This thesis examines how the introduction of generative artificial intelligence (AI) reshapes teachers’ work and professional learning, with a focus on negative consequences for workload, autonomy, and skill development. Drawing on a qualitative case study in a Bulgarian after-school English as a Foreign Language (EFL) center, the study analyzes three group interviews, each group consisting of 6 teachers, who use AI tools for lesson planning, feedback writing, and administrative tasks. Results suggest automation does little to reduce overall labor – rather, energy shifts toward checking outputs, adjusting inputs, rewriting prompts, tailoring results, while performance standards climb, increasing strain over time. Participants note subtle pressure to treat machine use as standard, along with unease about preset formats limiting classroom creativity and weakening teacher decision-making. Still, workers point to growing concerns about skill loss as everyday tasks move from creating content to overseeing outputs made by machines. Because of this shift, questions emerge around ethics – especially who is responsible when errors occur, how decisions are tracked, and whether monitoring becomes too invasive. The findings indicate that fair and lasting integration of AI into jobs relies more on workplace structures than on the tools themselves. Conditions like defined policies, space for teams to learn together, openness about how data is used, and strong safeguards for professional judgment in teaching contexts shape outcomes far more than code or algorithms ever could.