In a seminal study in 1996 , Kim et, al [1] demonstrated that a fusion of the C-terminal endonuclease domain of the FokI protein to a single zinc finger protein induces DNA cuts (ZFNs). This also applied to transcription activator-like effector proteins (TALEN). Later studies suggested that FokI needs to dimerize in order to cleave, without ever repeating the simple experiment mentioned above. Dimerization might increase efficiency, but the single monomer of FokI contains the catalytic site to cleave. If the monomeric FokI cleaves, the off-target problems with ZFN/TALENS would be significant in a large genome (especially with mismatches), since each component (left and right ZFN/TALEN) are typically 18 nucleotides or less. Previously, I had shown an unreported off-target in hornless TALEN-edited cattle [2]. Here, I show multiple off-targets in an ZFN study that edited two loci (AAVS1 and IL2RG) in stem cells [3]. The plasmid integrates at a loci in the HBB gene, which was not edited, across multiple(11) samples, with significant number of reads corroborating this fact. This also highlights the problem of plasmid integration in such gene-therapies, including the bacterial nuclease, which ought to be unacceptable. I also provide information on pre-clinical studies using ZFN that are now in clinical trials. Thus, more studies are needed to demonstrate safety in ZFN/TALEN studies with respect to off-targets.