If AI systems are to be used for truly human decision-making in education, teachers will need better support for deciding upon educational interventions and strategies on an ethically informed basis. As indicated by a recent call by the AIED Society to focus on the FATE (Fairness, Accountability, Transparency, and Ethics) of AI in education, fundamental issues in this area remain to be explicated, and teachers’ perspectives need to be accounted for. The paper offers examples of how AI may serve to promote learning but at the cost of presenting limited or untruthful information to the student. For example, false information about a student’s current progress may motivate students to finish a task they would otherwise give up; hiding information from the student that is disclosed to the teacher may decrease students’ cognitive load while supporting the teacher’s strategic choices, and deceiving the student as to the actual nature of the task or interaction, such as when using virtual agents, can increase students’ efforts towards learning. Potential conflicts between such scenarios and basic values of FATE are discussed, and the basis for developing an “augmented ethics” system to support teachers’ decision-making is presented. © 2020, Springer Nature Switzerland AG.