Very recently, Karati et al. presented a new generalized CLSC (gCLSC) to provide the functions of digital signature and encryption to fulfill the authenticity and confidentiality for the resource-constrained Internet-of-Things (IoT) devices. Karati et al. claimed that the newly proposed gCLSC supports the property of public verifiability and security of an ideal signcryption under the strong Diffie–Hellman and bilinear Diffie–Hellman inversion problems without the random oracle model. Unfortunately, their scheme is insecure with respect to unforgeability. It is demonstrated that any receiver or nonprivileged user can generate a valid signature or signcrypted ciphertext on any message. Finally, this article also points out the flaw in their unforgeability proof and gives some suggestions on the construction of a concrete scheme.