It looks like I was too mired in finals to catch
this story when it came out,
but I just learned of it today. Around
last November I was clued into issues surrounding Google’s autocomplete by this
story which prompted me to write an essay on whether similar litigation would succeed
in the United States. The essay is
forthcoming in the University of Illinois Journal of Law, Technology &
Policy, but I have not put it on SSRN just yet.
As I mention in the essay, lawsuits against
websites and other internet content providers are often futile when the
offending content has been provided by a third party. This is due to §230 of the Communications
Decency Act which states that internet content providers are not to be treated
as publishers of information that is provided by third parties. Courts have interpreted this provision to
broadly immunize websites from lawsuits arising from content provided by third
parties. For this reason, there is an
added obstacle to lawsuits against Google in the United States that may not be
present in other countries. This is
because autocomplete results are formed by algorithms that monitor popular
searches and suggest terms that are commonly associated with names or
topics. These users doing the searches may be considered third party content providers from whose conduct Google is immunized. I argue in the essay that this
obstacle may be overcome in the case of autocomplete defamation, however, since
the information that is posted through autocomplete is not provided for
purposes of posting on the internet, it is derived from individual searches
that people do not expect to be broadcast to the world.
Courts are generally leery of imposing liability
on websites because courts are worried that liability will chill internet
speech. Additionally, a concern that is
particularly relevant for websites like Google is that the breadth of content
they are dealing with is massive. There
is a strong argument that Google cannot be expected to monitor the output of
its autocomplete algorithm with respect to all of its search terms. This argument is weaker, however, if Google
is required to remove offending content only upon notification by the person
defamed. This seems to be what the
German court is requiring.
Successful lawsuits like the one in Germany are
relevant to potential lawsuits in the United States because if Google complies
with the resulting court orders, this compliance may serve as a model for how
Google may regulate its content in the United States if plaintiffs successfully
sue for defamation. While theoretical
arguments about chilling speech and restricting features may have persuasive
impact, situations like the one in Germany may provide an opportunity to see
the actual effects of restrictions on autocomplete, and show whether these
theoretical concerns will play out in the real world.
No comments:
Post a Comment