False positives are important, but they are far from the only thing.

First, a society that is aware it is under surveillance feels less free and people act less free, subjecting themselves to self censorship. See [[b:watched.html|A Watched Populace Never Boils]].

Secondly, effectively all surveillance technologies get repurposed and used for new activities, frequently pernicious ones. You have to in fact make the presumption of misuse. This applies not just in the country that first implements the technology.

You must also consider that once developed, it will be exported to countries with even worse civil rights records. (Example: U.S. Law requires Cisco to put wiretapping code into routers. Routers are then exported to China, Saudi Arabia, and even Greece.) Do we have any moral duty to consider what will happen if the good guys build surveillance infrastructure and then the bad guys find it easier to obtain or deploy?

There is only one solution, history teaches us. Minimal development and deployment.

The country with the best civil rights protections sets the bar for the world. It's well know that when the US, considered a bastion of free speech, adopts a restriction on free speech, the rest of the world starts arguing, "Well, if the USA can do this, and they're nuts about free speech, surely we can too."

Early on we asked the question, "Should the government have a big DNA database of anybody, so that if they get a DNA sample they can find who it is?" and so far the answer has been "no." Does the technology to spot relatives, which effectively gives us the quivalent of a DNA database of anybody, change our answer to that question?