#98527, By Moonbender crack down on kazaa???

  • Moonbender 28 Jun 2003 17:30:13 407 posts
    Seen 4 years ago
    Registered 13 years ago
    The only thing that nowadays is nontrivial to automate is the identification of copyrighted material (as you point out correctly). I assume a human will be necessary to do the final positive ID in every single case of alleged copyright violation. That's not to say the RIAA didn't fail to do this final step in the past, resulting in some rather embarrasing screw ups. I doubt they'll want to relive that kind of situation, especially now with the media spotlight on them. But that's not really a problem is it. How long does it take to positively identify a piece of copyrighted music? A minute is enough to listen to a third of a typical RIAA song, heck, 10 seconds should be enough to assure it is one of Spears' wailings and not free music renamed by some idiot.

    So a semi-automated process - note that I never said it can be totally automated - would look like this:
    1) get a popular song from any of the RIAA's members artists as a reference, make a number of operators listen to it
    2) search Kazaa (or whatever) for the song
    3) download the most popular results (most popular since well-spread files are relatively more likely to be non-fake)
    4) while downloading, keep a tight log on what's happening: IP address of the other side(s), time and date, search keywords, filename, etc
    5) have the operators do an ID to verify whether the file is actually the one you were looking for, discard "fakes" - or save them for special treatment since they are almost guaranteed to infringe on RIAA copyright, too
    5) identify the ISPs owning the IPs you identified
    6) have form-letters sent to the ISPs demanding information on who used the logged IPs at the logged time in accordance with the DMCA Digital Millenium Copyright Act and the NET No Elecronic Theft act
    7) have form-letters sent to the people named by their ISPs calling them to cease and desist and (optionally) pay some sort of reparation or face court action

    That's it. In principle, every step apart from 5 can be automated, although they can just as easily be done manually as well, and I'm sure they'll take care not to hit any innocent folks in the beginning (not that there's a fat chance of accidently hitting the 20% or whatever internet users that have not on occasion commited a copyright violation).
    Most of this has been possible for a while, too, all but number 6, forcing the ISP to hand over personal information on their customers based on no evidence, without a court order.
Log in or register to reply