Framed by Google
My FindLaw column for today (available here) discusses a recent Ninth Circuit ruling (available here) reversing a preliminary injunction against Google's image search engine on the ground that its thumbnail images infringed the copyright of the owners of the originals of those images. The 9th circuit found that the thumbnails were fair use, at least absent a specific showing that their economic impact on the owner's business---a porn website which also delivered low-res pics to cell phones---outweighed the transformative nature of the thumbnails when used as part of the image search on a conventional computer. The court also found no prima facie case of infringement from Google's "framing" of images from infringing websites, where those images resided on the infringers' servers rather than Google's. In other words, it applied what the district court called the "server test."
If that went by too quickly, read the column, which explains these issues in much greater detail. Actually, read the column no matter what. It's right here. Go ahead, read it. I'll wait.
Okay, glad to have you back. Anyway, as you now know from reading the column, I suggest there that the server test makes it too easy for copyright scofflaws to appropriate images without running afoul of the law, simply by embedding links to licensed images on other sites. My colleague Tim Wu points out that this is true but also agreed with my suggestion (not made in the column but available exclusively here on Dorf on Law!!!) that there ought to be a technological way to avoid this problem. If you don't want Google to index and/or cache your site, you just need to include a line of code so stating. As I discussed a while back (in a post here), a Belgian court has ruled that this opt-out system insufficiently protects copyrighted material, but Tim disagrees. He notes that the vast majority of site operators want to be findable by Google, and so a Googlable default makes sense.
That works if you're worried about Google but what if you're worried about every tom, dick and harry with a blog borrowing your content? One possibility would be a legal rule that says that use of the Google-don't-cache-my-page code also makes it illegal for anybody else to frame your content without specific permission, even if they don't copy your material onto their server. Absent such a rule, owners of copyrighted material may resort to self-help, embedding encryption in images and other content to prevent them from displaying when framed by non-licensed sites.
If that went by too quickly, read the column, which explains these issues in much greater detail. Actually, read the column no matter what. It's right here. Go ahead, read it. I'll wait.
Okay, glad to have you back. Anyway, as you now know from reading the column, I suggest there that the server test makes it too easy for copyright scofflaws to appropriate images without running afoul of the law, simply by embedding links to licensed images on other sites. My colleague Tim Wu points out that this is true but also agreed with my suggestion (not made in the column but available exclusively here on Dorf on Law!!!) that there ought to be a technological way to avoid this problem. If you don't want Google to index and/or cache your site, you just need to include a line of code so stating. As I discussed a while back (in a post here), a Belgian court has ruled that this opt-out system insufficiently protects copyrighted material, but Tim disagrees. He notes that the vast majority of site operators want to be findable by Google, and so a Googlable default makes sense.
That works if you're worried about Google but what if you're worried about every tom, dick and harry with a blog borrowing your content? One possibility would be a legal rule that says that use of the Google-don't-cache-my-page code also makes it illegal for anybody else to frame your content without specific permission, even if they don't copy your material onto their server. Absent such a rule, owners of copyrighted material may resort to self-help, embedding encryption in images and other content to prevent them from displaying when framed by non-licensed sites.
posted by Michael C. Dorf @ 7:29 AM
2 Comments:
At 9:25 AM,
egarber said…
It seems to me that Congress would be the better source for a standard solution. Otherwise, the courts would have to create remedies far beyond their expertise -- and there might be 50 competing conclusions.
Of course, I don't want Congress choosing the exact technology, but an overall specific mandate (similar to the one requiring a move to digital television signals) could work, I think.
Do the existing copyright laws mandate something like this to any meaningful degree?
At 10:50 AM,
Ben said…
There's an additional wrinkle here.
As you say in your article, a simple and clever way around copyright restrictions on an online image is to hotlink directly to the image on its original server. However, many bloggers who post images on their sites don't like this kind of direct hotlinks. The reason involves server usage.
Let's say I want to post an image I've found on a small blog on a very high traffic blog, like Daily Kos. If I embed an image by hotlinking directly to the small blog, I'm going to push the small blogger's server use through the roof.
On the other hand, if I upload the image to an image-hosting site, like ImageShack, I'm not overstressing the small blogger's bandwidth. I am, however, potentially running afoul of copyright laws in a way that I wouldn't be had I just hotlinked to the image on the small blog itself.
Post a Comment
<< Home