The victim of this policy was not the (hypothetical, I have no idea about the actual case) person who was both the victim of a crime and perpetrator of another

The victim of this policy was not the (hypothetical, I have no idea about the actual case) person who was both the victim of a crime and perpetrator of another crime, it's the much larger number of people who did not commit a crime but are being "investigated" every time a suspect's DNA is run through the system because they reported being raped. They are subjected to the possibility of false positives or misleading true positives (it is their DNA but they didn't do the crime). These are low probability events but the population is fairly large and the number of suspect samples is also large and one images both will only grow in the future. But I think the real problem is that it shows a certain lack of thoughtfulness when it comes to data use and governance. It doesn't seem like they considered themselves stewards of the most sensitive data possible from people who have suffered horribly. It's possible they weren't aware of the ethical quandary here or they didn't have the technological sophistication to develop a nuanced solution. Either case argues for more restriction in how the data are used while we as a society catch up to the implications of this evolving technology.