Recently, Bitansky [Bit17] and Goyal et. al [GHKW17] gave generic constructions of selec-tively secure verifiable random functions (VRFs) from non-interactive witness indistinguishable proofs (NIWI) and injective one way functions. In this short note, we give an alternate construc-tion of selectively secure VRFs based on the same assumptions as an application of the recently introduced notion of verifiable functional encryption [BGJS16]. Our construction and proof is much simpler than the ones in [Bit17, GHKW17], given previous work (most notably given the constructions of verifiable functional encryption in [BGJS16]).