A number of web companies are teaming up to create a database of child porn images, which they claim will be used to monitor for child porn being distributed around their networks, in much the same way antivirus companies keep virus databases, and antispam companies keep corpuses.
I’m afraid the analogy falls down for me there. Anyone that’s ever worked with digital images knows that as soon as you resize the image, the file becomes completely different. The filesize changes, the datastamp changes, and the content of the file changes. Sure, there are algos for comparing images based on how they “look”, but aren’t these very hit and miss?
I’d be interested in the opinions of digital artists and graphic designers, because as far as I can see this looks like an initiative created to make it look like something is being done, instead of actually doing something.
Plus of course the global move to broadband and the massive growth of P2P creates other questions, such as: Who looks at porn jpegs anymore, for chrissake? JPEG is sooo nineteen nineties!
(I’m being facetious, perhaps these scumbags are restricted to JPEG because of difficulties with distribution. Still, it seems unlikely.)