Transposable elements can alter gene regulation in their host genome, either when they integrate into a genome, or when they accrue mutations after integration. However, the extent to which transposons can alter gene expression, as well as the necessary mutational steps, are not well characterized. Here we study the gene regulatory potential of the prominent IS3 family of transposable elements in E.coli. We started with 10 sequences from the ends of 5 IS3 sequences, created 18,537 random mutations in them, and measured their promoter activity using a massively parallel reporter assay. All 10 sequences could evolve de-novo promoter activity from single point mutations. De-novo promoters mostly emerge from existing proto-promoter sequences when mutations create new -10 boxes downstream of preexisting -35 boxes. The ends of IS3s harbor ~1.5 times as many such proto-promoter sequences than the E. coli genome. We also estimate that at least 26% of the 706 characterized IS3s already encode promoters. Our study shows that transposable elements can have a high latent cis-regulatory potential. Our observations can help to explain why mobile DNA may persist in prokaryotic genomes. They also underline the potential use of transposable elements as a substrate for evolving new gene expression.