I won't be surprised if Bush tries it out as his next lame rationale for the war on Iraq. For now we'll settle for the RIAA saying that it's not just trying to stop music sharing it wants to protect the kids. A pedophile could send "an instant message to the unwitting young person who downloads an Olsen twins or Pokemon file from the pedophile's share folder on Kazaa," RIAA chief Cary Sherman said, sexing up, just a bit, a GAO report. (It said that some files with those keywords were actually porn -- more on that shortly -- not that pedophiles were IM'ing people who searched for them).
Now for a couple of oddities within the GAO report. Here's the The New York Times paraphrase:
A study in March by the General Accounting Office found that KaZaA would be effective for someone looking for child pornography. The agency searched for 12 terms associated with child pornography, such as "incest" and "underage." It did not actually download the files it found, but it determined that 42 percent of them had titles or descriptions associated with pornographic images of children.
So, my first reaction was, They didn't download the files? That means they didn't learn anything, since everyone knows that porn-spammers spice up their files names with keywords that have nothing to do with the actual images. Then I re-read the paragraph and realized that what the GAO was saying is that even when you go out of your way to look for kiddy porn, less than half of the results will even have "titles or descriptions associated with pornographic images of children," much less actually BE pornographic images of children. WTF? I mean, when you use "incest" or "underage" as a search term, shouldn't 100 percent of the results contain those search terms, even if the file name is ultimately inaccurate? Unless the Times just got this wrong, KaZaA's search function is seriously flawed (or are there that many non-porn images that are accurately described with such words? If so, what would they be of?).
Now an interlude from another paraphrase of this report:
The GAO's auditors chose not to open them because under federal law, it is illegal to knowingly possess child pornography...The auditors did, however, ask the U.S. Customs' CyberSmuggling Center to test a smaller number of images found using three keywords related to child pornography. "The CyberSmuggling Center analysis of the 341 downloaded images showed that 149 (about 44 percent) of the downloaded images contained child pornography," the report says. "The center classified the remaining images as child erotica (13 percent), adult pornography (29 percent), or non-pornographic (14 percent)."
That seems to be a significantly higher percentage of ACTUAL kid-porn than the GAO's own test would have turned up, had they followed through. BTW, I wonder if, say, Trent Lott knows that the U.S. Customs office is being paid to parse the difference between "child pornography" and "child erotica."
Back to the Times:
A second aspect of its study measured the likelihood a child would inadvertently be exposed to pornography using KaZaA. It examined 157 files downloaded in response to three search terms of interest to children Britney, Pokémon and Olsen twins. It classified 49 percent of those files as pornographic.
OK, that's a little creepy, though I'm curious how those 77 hits break down (Britney 72, Olsen twins 5, Pokémon, 0?). That seems like pertinent information. Also, the paraphrase makes it sound like kids looking for Britney Spears MP3s might stumble onto porn, when it's likely that the GAO singled out image files (or it would have had a much lower "success rate") (Update:Daze points out that MSNBC states outright that hardcore porn is "finding its way into music files.") Anyway, here's News.com again with the CyberSmuggling Center's similar test:
Of the 177 images the CyberSmuggling Center downloaded from Kazaa using "three keywords representing the names of a popular female singer, child actors and a cartoon character," it classified only two as falling into the category of child pornography. The remainder would be legal to possess--and legal to distribute assuming they did not violate other restrictions such as obscenity or copyright laws..
Only two? That's a huge difference (and it's one reason you won't see the C3 version of this test reported very often). Update: Ah, I wasn't reading closely enough. The Times (following the GAO's lead?) switched gears without warning, jumping from child pornography to the ordinary variety in successive paragraphs. I blame the media for my confusion.
In summary, if you're looking for fake nudie pics of Mary Kate and Ashley, the GAO is gonna do a much better job of finding them for you than the CyberSmuggling Center. I wonder how much that study cost....