It depends on how we make the enzyme. If it's made without involving life in any way, it's hard to imagine how bacteria would "learn" to make it after finding i

It depends on how we make the enzyme. If it's made without involving life in any way, it's hard to imagine how bacteria would "learn" to make it after finding it. At most, opportunistic bacteria could better parasite people already loaded with enzymes from CRISPR treatment. If we modify harmless bacteria to build the enzyme for us, then sure, the modification could well end up on harmful bacteria.