This clearly is a new area in ICT and therefore ready answers would not be available. But as a community, we would need to watch the issues that arise and ust that to craft acceptable rules on the use of the technology.

The cons of the technology have been pointed out, and we "collectively" must work to ensure that its benefits are encouraged and its misuse discouraged.

Exactly the point.  There is a "life cycle" of information: birth (gathering), life (use, transmission) and (one hopes) eventually death (removal of information from databases, when no longer useful).  As we generate ever-greater memory and transmission capacity, data creation, use and retention tends to expand.  With greater instant access and coordination of such information, moreover, we may move closer to a world where privacy is very difficult to preserve.  Thus, even if we retain a sense of an "expectation of privacy," the technology could become overwhelming.

Again, I am not a Luddite.  I suggest only that we, as a society, think about what we are creating, and attempt to place reasonable controls, where appropriate, to protect essential notions of privacy.  How those controls are mostly effectively maintained (through private contracts, through "best practice" voluntary standard-setting, or through regulation) is part of the debate.

Steven!  The same question  is over and over again. as soon some new technology  appears, the question of privacy is raised or (some other ethical questions). The  new technologies will always raise a lot of ethical questions but ethics  will never stop a development of technologies(alas sometimes). so we can discuss what bad or good may happen, but it won't change anything

I feel lucky that we even have a need for reasonable expectation of privacy and demand so much evidence to identify a person as a suspect - Our rules make it at least a little difficult to take advantage of the software and software to come

When I read about China using their "Fast Human Face Recognition System" at the Beijing Olympics to screen over 80,000 people in one day via 100 checkpoints, I am more concerned with how governments such as those will take advantage of the future of facial recognition - with their lack of privacy laws

Part of the problem, in my humble opinion, is that we are raising a generation of on-line voyeurs and exhibitionists.  What Andy Warhol said would be "15 minutes of fame" for everyone has more or less come true.  Lots of people seem quite content to give up their privacy, in exchange for very little other than the thrill of being recognized.  What does that do to our "expectations of privacy," one of the touchstones of privacy law?   

Don't get me wrong.  I'm not a Luddite.  I can't see trying to stop this technology, any more than we could stop the development of trains, autos, telephones, etc.  Rather, the real key, again only in my humble opinion, is to think about these technologies, as a society, and decide how we wish to deal with them: 1.  Purely private, "every person for himself" remedies (perhaps we all wear hoodies and masks when we don't want to be recognized).  2.  Private, but associational remedies (establishment of "surveillance free" zones of property, by agreement).  3.  Development of the law, through case by case judicial pronouncement, on what constitutes extreme, unacceptable intrusion into privacy / seclusion.  4.  Creation of some general legislative mandate, and assignment of responsibility for enforcing the mandate to one or more administrative agencies.  5.  Creation of more detailed legislative codes to protect privacy.  6.  Enshrinement of privacy principles into the Constitution itself. 

There are two different questions here (in the legal context):

1.  Can we rely on facial recognition technology to identify perpetrators of crime (i.e., to serve as another valid law enforcement investigative tool).  For this question, the problem of "false positives" is a key.  If facial recognition technology too identifies innocent people (and especially if it too often fails to identify real suspects), it may not be considered a reliable source of information for law enforcement purposes.

2.  Can we rely on photographic evidence, which may be subject to alteration?  In the context of admission of evidence at trial, the standards essential revolve around "foundation" and "authenticity."  Can some witness attest that the evidence (here a photo) is what it purports to be?  And, is there any question whether the evidence (photo) is authentic (i.e., altered in some way)?  By and large, courts allow the admission of evidence with relatively low levels of assurance on both these questions, and permit the cross-examination process, and admission of contrary evidence to serve as a check on problems.

One can imagine, however, courts taking a more jaundiced view of facial recognition / photographic evidence, if there were some clear indications of abuse. 

See, by comparison, the debate over the use of "lie detector" technology as a tool in law enforcement, and as evidence in court. 

It's one of those things that when you hear about it, you say "cool" but then you start thinking that it has too many hurdles to surpass. What about people on witness protection? 

But it's definitely an interesting idea.

"Is it legal to take a picture of someone without their consent?"

3 points to consider:

1-  you can photograph private property (including people) as long as you’re not trespassing to get the shot.  

2- you can’t snap shots of people in the bathroom, a dressing room, or similar places. (You will be violating their privacy)

3- Don't take pictures where there is a question of national security.

Apart for that, there is no law against taking photos of somebody and upload them on a website. 

Very thought-provoking post, Steven.

One thing I believe we will need to address soon, is establishing a baseline for the courts of validation a picture/scan/facial recognition of a person for use in cases.

Right now, a good Photoshop person can craft something that can fool a judge or jury.  And with 3-D imaging technology (like that used in Avatar), you can literally put a person in any situation that would look incriminating.

If I were scanned, I could be (erroneously) associated with any number of nasty events through adjusted pictures. 

One thing I proposed a long time ago (back in the 80's) was integrating a checksum into pictures and images (primarily radiology imaging, back then) and with the use of digital imaging being so prevalent, it seems even more important to have a checksum that ensures that either the picture is not altered or the fact that it is altered is clearly evident.

Once you can establish that a picture (or video clip) is not doctored, then the issues of which facial recognition standards to use becomes a solveable problem, at least for humans.

Answering the question posed in the column,

"Is it legal to take a picture of someone without their consent?"

The Govt already does it without our consent(at various public places like Airports ,Metro Stations,etc,etc).

This just ensures that more corporations are able to do it (in the name of Technology or progress or whatever...)

Is it right or is it wrong? It certainly has frightening privacy implications for sure.But lets leave that question for the courts to decide how they should implement this in our society which on the one hand is increasingly paranoid about privacy concerns and on the other hand wants to give up this privacy in the name of feeling safe under a nanny state.I dunno which is better ....

Regards

Ashish.