As a result of the cancer immunotherapy revolution, more than 2000 immuno‐oncology agents are currently being tested or are in use to improve responses. Not unexpectedly, the 2018 Nobel Prize in Physiology or Medicine was awarded to James P. Allison and Tasuku Honjo for their development of cancer therapy by the blockade of co‐inhibitory signals. Unfortunately, manipulation of the co‐inhibitory receptors has also resulted in a safety issue: widespread iatrogenic immune‐related adverse events (irAEs). Autoimmunity is emerging as the nemesis of immunotherapy. Originally, it was assumed that CTLA‐4 blockade selectively targets T cells relevant to the antitumour immune response. However, an uncontrolled pan T cell activation was induced compromising tolerance to healthy self‐tissues. The irAEs are very similar to that of a chronic graft‐versus‐host‐disease (GVHD) reaction following allogeneic bone marrow transplantation (BMT). We hypothesized that ipilimumab induced a graft‐versus‐malignancy (GVM) effect, which eradicated metastatic melanoma in a minority of patients, but also involved an auto‐GVHD reaction that resulted in widespread autoimmunity in the majority. Therefore, we argued for a profound theoretical point against the consensus of experts. The task is not to desperately put the genie back in the bottle by immune‐suppressive treatments, but instead to harness the autoimmune forces. In this way, the same goal could be achieved by an antibody as by the adoptive transfer of alloreactive donor lymphocytes, but without severe GVHD. The proof‐of‐principle of a low‐dose‐combination immune checkpoint therapy, consisting only of approved drugs and treatments, was demonstrated in 111 stage IV cancer patients.