Comment by icegreentea | Hacker News Reader

Comment by icegreentea | Hacker News Reader

Parent Full thread

Potentially no one is lying, because even if the information in the slides were 100% accurate, they could describe a wide range of possible programs and implementations, many of which are also 100% inline with the companies' claims.

For example, if PRISM is really just an internal NSA tool for interfacing, searching, storing etc information procured from these providers from different means (such as NSLs or FISA warrants), then obviously no company joined it. It's an internal tool! The idea of 'adding' just becomes ensuring normalization of data and other such 'mundane' things.

Note I say, 'for example'. Because amongst all the yelling about, we still don't actually know what the fuck PRISM actually is, and arguments just revolve around how charitable/cynical you are about the parties involved.

The 20M project could well be a number of things not involving the companies direct participation

* An interface to fuse data received by court order from each site.

* A scraping tool to pull in open source data, possibly with seed accounts (i.e., a facebook account for each university) to give you better access.

* Scraping data on the wire to and from each service, at the level of the internet backbone

This is also what I sort of assume. All of these tech companies like facebook, and google, probably got so many requests for info from the feds that at a certain point, it must have somehow made sense to switch it up and just give them an automated web form somewhere. (direct access to servers..) The telcos have probably had these since way back.. Eventually, there would be quite a few of these web forms, so some 'top' coders put together a system for the feds to query all of the forms at once with one request. Paint on a GUI and call it PRISM. It was nice to hear that Yahoo! still sorts out their fed requests by hand.

Facebook and Google (Zuck & Page) have also stated that each request for data is inspected individually. I don't know where you're getting that there's some automated system for pulling user data.

Reply on news.ycombinator.com