For example consider the case where a couple with fertility problems used a sperm donor. The child is never told that this happened. Then police identify that the crime was likely committed by someone who shares 1/4 of their DNA with the child. In other words that child's grandchild, grandparent, half-sibling, aunt or uncle. You can imagine the sudden disruption when the police show up?

It gets even worse if the child was the result of an affair. Or if the person who committed the crime was. Or if the child was in some other way the result of a family secret.

Oh right, and it gets more problematic. The traced DNA is not necessarily the criminal's. You find hair at the crime scene, but maybe it was present before the crime. The hair is a lead, it is not proof.

These cases are common enough that schools long ago learned not to use people's families for demonstration of how genes work in biology classes. Heck, they are common enough that I personally know people that I know are in these kinds of situations. And you have to get to know someone pretty well before you find out about stuff like this.

And now we're going to have law enforcement randomly exposing private situations like this. Maybe we as a society consider this worthwhile. But there is no question that privacy will be invaded here.

If the police see a red BMW leave a scene, it might accidentally cause a person with a red BMW to be falsely investigated.

Human eyewitness evidence is even worse.

So yea, while there is some effect, it is no more than what we already accept.

Additionally, your hypothetical also assumes that the person(s) related to the supposed subject have their DNA on file or are somehow compelled to provide it as evidence - precluding some kind of existing database, or a kind of chicken|egg situation. That is out of scope of this argument.

That is actually a very safe assumption.

As articles like http://www.pbs.org/wgbh/nova/next/body/dna-databases/ make clear, police already ARE creating DNA databases. And since there is little regulation around this right now, they are free to be creative in assembling them. For example you can get a good DNA sample from the straw used to blow into a breathalyzer. So get stopped at a routine traffic stop, your name, license plate, and DNA can go into a database. And voila, police can now match you and your relatives against crime scenes!

Think this is crazy? This has already been happening. See https://www.prisonlegalnews.org/news/2013/aug/15/maryland-co... for breathalyzer DNA being used to convict of a crime, http://edition.cnn.com/2013/02/26/justice/supreme-court-dna/ for the Supreme Court OKing DNA tests for arrests without a conviction, and http://www.nbcdfw.com/news/local/North-Texas-Drivers-Stopped... for a random example where this was done to random people at a traffic stop. Yes, there is outrage at present. But over time this could become the new normal.

As technology improves and costs come down, incentives to match everyone to DNA tests will go up, not down. Heck, I remember as a child in Canada over 30 years ago having police come to my elementary school and taking everyone's fingerprints. I believe that the cause was so that if anyone went missing, they could identify us. But once that data is collected and goes into databases, it doesn't come out.

Heck, how many parents would give consent today for their 2nd graders to get a DNA test for that purpose?

Won't let me reply next-in-thread. I never said that what you describe wasn't possible, and I'm indeed aware of those instances you bring up. What I'm saying is that that is out of scope of this argument. EFF is talking about collection capabilities. Immediately jumping to the 1984 senarios hurt the overall grounding of the argument.

Therefore if there is a demonstrable privacy issue to DNA collection and analysis, then requiring warrants is the accepted way in our legal system to balance the needs of law enforcement against people's right to privacy.

Or your child. I guess you don't care if you send your kid to the chair though.