This theoretical framework draws on my published research to take a critical approach to Artificial Intelligence (AI) in education. It comprises four defining features that are designed to raise awareness and provoke discussion amongst teachers, students, institutional policy makers and educational technologists around the nature and purpose of teaching in the age of AI. The framework seeks to embrace a more values-based concept of what good teaching should look like. Each of the features is accompanied by a question that functions as a Call to Action. The questions are designed to provoke human educators to consider how they might work with AI in education in ways that are educationally desirable. Political Values: To what extent and on what terms is your teaching directed and controlled by AI? (Gibson, 2024) AI driven educational technologies have the power to control human behaviour through their ownership and analysis of the huge volumes of user-generated data. This datafication of education is of a cyclical nature. New data-driven theories of learning emerge, which are in turn coded into the educational technologies, academic institutions purchase these technologies, pedagogic practices are influenced by the use of these technologies, data-driven insights produce the evidence to create policy, which, in turn, feeds back into and validates the learning theories. Furthermore, this over-reliance on factual information to create evidence-based educational policies and practices erodes human judgements around what is educationally desirable. Ethical Values: How might you generate student awareness around the algorithmic agency of AI? (Gibson, 2023) AI is an active agent in shaping our knowledge systems. However, AI hallucinates and is prone to bias and prejudice. We must push back against the encroaching political forces of AI monopolising our educational practices. Freedom in the future is contingent to our understanding of the technologies we use, how they work, why they work as they do, who controls them and what value systems they embody. This is an ethical response in its aim to generate understanding around the political forces within our relational encounters with AI. Pedagogical Values: How might your educational activities be developed to be more collaborative, rather than individualistic? (Gibson, 2025) In education, we tend to conceptualise AI as a personalised one-to-one private tutor whose primary function is to facilitate the transmission of knowledge in an efficient and scalable manner with little consideration for the messiness and unpredictability of the lived experience of teaching with AI. This individualistic approach positions AI as a neutral tool and the human teacher and students as passive objects. To embrace a more collaborative approach is to position all entities as active subjects. In this capacity, knowledge is generated in and through the relational encounters. Social Values: How might you and your students become involved in the design and development of your AI technologies? (Gibson, 2023) Teachers should become more involved in shaping their educational technologies to become more pedagogically driven, reflective of our value systems, and respectful of the democratic boundaries between humans and technology. Human teachers might work with computer programmers, philosophers, political theorists, designers, and sociologists at an institutional level.