Hmm, I wonder if you can be legally held liable for the work of an algorithm if no malicious, slanderous, or personal intent could be proven. In other words you code the algorithm without bias to site or author, but by clearly weeding out a class of content/links you see as less "relevant". It's just another search engine. It never targets anyone unduly - it's just part of its algorithm. A new site could arise that perfectly plays against your algorithm's coding, and gets essentially buried in the search results, but you didn't code it knowing of that site, so how is it discriminatory or illegal? I don't know, I'm not a lawyer, and some of the things people sue over these days drive me nuts (what am I saying "some" - MOST). But it seems reasonably defensible...
- Oshyan