Despite recent advances, the prognosis of acute myeloid leukemia (AML) remains unsatisfactory due to disease recurrence and the development of resistance to both conventional and novel therapies. Engineered T cells expressing chimeric antigen receptors (CARs) on their cellular surface represent one of the most promising anticancer agents. CAR-T cells are increasingly used in patients with B cell malignancies, with remarkable clinical results despite some immune-related toxicities. However, at present, the role of CAR-T cells in myeloid neoplasms, including AML, is extremely limited, as specific molecular targets for immune cells are generally lacking on AML blasts. Besides the paucity of dispensable targets, as myeloid antigens are often co-expressed on normal hematopoietic stem and progenitor cells with potentially intolerable myeloablation, the AML microenvironment is hostile to T cell proliferation due to inhibitory soluble factors. In addition, the rapidly progressive nature of the disease further complicates the use of CAR-T in AML. This review discusses the current state of CAR-T cell therapy in AML, including the still scanty clinical evidence and the potential approaches to overcome its limitations, including genetic modifications and combinatorial strategies, to make CAR-T cell therapy an effective option for AML patients.