Obviously we can construct an agent who does this. I just don't see a reasonably parsimonious model that does it without including a preference for getting AIDS, or something similarly crazy. Perhaps I'm just stuck.

Are you talking about the idea of a Giffen good, where bread becomes more costly, reducing your household budget, forcing you to get more of your calories from cheaper goods such as bread, meaning you buy more bread? If so I'm not sure I see how that applies in this instance.

In general I don't mean to make an argument that it's impossible for harm reduction strategies to be counterproductive, but that condoms in particular are so very effective at reducing HIV transmission that it's simply implausible to posit that any countervailing effect could do so much as to make the overall outcome worse - such an idea would need very strong evidence, which plainly isn't here.