Monthly Archive: April 2012

the (legal) future of collaborative documents

Fascinating post about the US v. Lawson case over at the blog run by Eric Goldman (a professor here at SCU Law).  It’s actually a guest post by Venkat Balasubramani.

I’m not a lawyer, do not have a JD, and am generally often only about 95% informed on things which makes me very dangerous and likely to make myself look foolish.  To make things worse, the blog belongs to a professor at the law school where I work.  I don’t know the guest author, but this is still a pretty dicey situation.  But I’m going through with this post anyway.

Basically, the ruling in US v. Lawson, which was essentially about cockfighting (or “gamefowl derbies”) but involved issues of “sponsorship” of the events, was overturned because a juror printed up the definition of the word “sponsor” off of Wikipedia and brought it into deliberations.  That, unto itself, apparently would not necessarily warrant overturning the decision – the original court ruled that it did not prejudice the jury.  The Fourth Circuit, on appeal, ruled that it did prejudice the jury, however, and issued an opinion (but not a decision on the appeal?  this is where my understanding of the various rulings a court can have gets really muddy) on the matter.  The opinion is fundamentally arguing that the inherent effect of bringing an outside source into the jury deliberation room.  It also, however, makes a lot of comments about wikipedia specifically, and suggests certain things about other collaborative editing situations.

One way the court discusses the possible validity of the document brought into the jury room is whether the definition procured from Wikipedia was the legally correct one.  But since Wikipedia is always changing, someone would need to prove that the definition on wikipedia the day that the juror printed it out was accurate, not just the one that is on the site “today.”  The government didn’t bother to do that, which is weird but that isn’t want struck me.

What really intrigues and disturbs me is this quote from the ruling:

the court notes that even if historical edits were presented by the government, it could not consider these, absent some indication that Wikipedia archives of historical changes are “accurate and trustworthy.

So…while I think it’s awfully strange that the government didn’t even try to retrace the versions of the wikipedia page, the court would need an “indication” that the archives would even be accurate.  The court wouldn’t even consider the evidence presented unless it could be “proven” that Wikipedia archives were accurate.

What would be sufficient indication?  What would be enough to make one court feel that the archives of Wikipedia (or…revision history in Google Docs?  on Dropbox?) are accurate and reliable?  Wikipedia is based on the MediaWiki platform but what about other ones (like…Wordpress, which is the one that powers this blog?)  Would that “indication” for the Fourth Circuit be enough for the next court if there is an appeal?  What about a court in one state vs. the next?

Considering how many different collaborative editing tools we use these days, this is, to me, a frightening question.  Almost all of these tools – whether a productivity tool like Google Docs or a file storage/sharing system like Dropbox or Box – have some kind of version history system.  If for some reason a document that has been edited collaboratively becomes important in a legal case, what if the court decides that Box’s version history system is not sufficiently accurate?  That it is possible to somehow “break into” the history and alter it, such that it is no longer a reliable part of an e-discovery effort?  Or what if one person collaborating on the document – a “low-level” team member – went through the revision history and restored the document to some earlier point?  Then there is more editing.  If you go from version 1 to version 3, then back to version 2, then edit to version 4…I have no idea.

Do we now have to start vetting archiving and versioning systems to meet exposure, e-discover and other legal needs in addition to just general security of information?  And should we just be paranoid in general?