Most of the organizations that are coming out against TIA are, in my opinion, way off base -- due mostly to an inate lack of understanding of how the US Intelligence apparatus operates, and what strictures it operates under. This is doubly amusing, since most people railing against TIA are the relentless technophilic Wired-magazine-reading optimists that typified the late 90s.
First, you have to realize that there are virtually no limits to what data can be collected about non-US persons. This is a fact of life; it is definitely a jumping-off point for debate, however -- you may argue that it is wrong, immoral, or otherwise bad to do this (although I believe this flies in the face of evidence of 200+years of intelligence tradecraft). However, this argument is a non-starter when it comes to TIA -- the Intelligence agencies ALREADY collect this data, and they will continue to do so, TIA or no TIA.
Now, during this collection process, they are bound to slurp up prohibited information, such as data about US persons. The IC operates under very tight rules about this: the law forbids collection and identification of information about US persons, and any information about them must be immediately redacted from the collection, almost ad absurdum -- for instance, see this link:
http://www.epic.org/privacy/nsa/documents.htmlOf course, people rightly ask "but won't TIA circumvent this?" The possibly exists. Note that I said "the law forbids", and as anyone knows, laws can be broken. But that is true with or without the presence of TIA. Technology that allows you to sift, filter, analyze, manipulate, or search data is useful regardless of the source of data you put into it. If you are the CIA or NSA, and put in US-sourced data, you are committing a crime. But don't delude yourself into thinking TIA is a necessary factor there.
The only thing that will prevent systems like TIA (or any system the IC has now, or any system it might have in the future) is careful crafting of rules, and monitoring to ensure compliance. Railing against specific proposals or configurations is a mindless waste of time. I forgot to mention that I am one of those technophilic Wired-magazine-reading optimists, and believe me, this technology is coming, like it or not. In 15 years, the IC may be able to go down to CompUSA and buy products that do what they are testing now, developed by some commercial source for some purpose which will seem normal 15 years hence (and believe me, they'd prefer to be able to do that!).
The problem that all the technologies underlying TIA are trying to solve are very, very basic ones in information management: making structured data out of unstructured information, and being able to reason based on the structure you have created. This is in many ways a holy grail for anyone doing research in information management, and it's 100s of times harder than what Google seeks to accomplish. Tou can imagine the types of people that could use it: for instance, an investment banker that has all newscasts regarding a particular topic or company transcribed and translated, then linked showing cross-board and cross-investment linkage looking for buying or selling opportunities.
I would suggest that people who are concerned about TIA adopt a different set of tactics than running around crying that their privacy is being invaded.
1) Write to your congresspeople asking for greater oversight of the intelligence community;
2) Read all the published papers that come out of research areas that would be of interest in a TIA-like program (surprise: 99.999% of it is unclassified and published in conference literature and journals), and work to build your own system, and understand how to make things like this. After all, "Scientia Est Potentia".
