« Does Google Cloak? | Main | Shopbots and Nash Equilibrium »
March 08, 2005
Google AutoLink and Copyright Law--A Fuller Explanation
There’s been a lot of discussion about Google’s AutoLink tool, including some questions about why I think it might violate copyright law. I can understand why this is confusing, particularly because the AutoLink tool raises two discrete questions. First, what is the best policy result? Second, what does the law say today? We need to answer these questions separately.
On #1, I'm in favor of technology that improves people's lives, such as by saving them search time. Therefore, I want to find a way for tools like AutoLink to exist.
On #2, I think AutoLink may create a derivative work. Google modifies the screen display by layering its own content on top of publisher content. The fact that users may ask Google to modify the screen display doesn't change the analysis.
To understand why Google is a legally significant actor, we should distinguish the following three situations: (a) Google acts on its own, (b) Google acts as an agent for the user, and (c) a user acts on his/her own. Consider the latter: if Google gave the user clip art and the user pasted the clip art on top of a work in a way that created a derivative work, then Google isn't responsible for the derivative work. In contrast, if a user asks Google to create a derivative work and Google does so, then at minimum Google created a derivative work (and maybe the user is liable as well). In my mind, this describes AutoLink.
For a precedent, consider UMG v. MP3.com, when MP3.com launched the my.mp3.com service. The service let users listen to music they already owned. But to let users do so, MP3.com committed infringing acts, and MP3.com couldn't simply claim to be the user’s agent.
My legal observations are positive, not normative. I'd be delighted to be wrong on the legal question. But we have to answer the positive question based on the law today, not as we wish it were.
Posted by Eric at March 8, 2005 06:26 PM | Copyright