[ NNSquad ] Re: Liability issues in ISP-injected ad systems?

[David Ulevitch <david@opendns.com>]

As we go down this inevitable path it is important we don't  
simultaneously do something that forces us to accept spam or other  
unwanted traffic at the edge of the network. It may seem like a  
different issue but it is closely related. See arguments in comcast vs  
360 or vixie/MAPS history for the other side -- the "good" kind of  
packet molestation that can be impacted.

Sent from my iPhone

On Apr 20, 2008, at 9:32 AM, Lauren Weinstein <lauren@vortex.com> wrote:

> In discussions of ISP injections of messages and ads into
> third-party data streams (Rogers has now gone live with injected
> messages in Canada), most of the analysis so far has been focused on
> copyright issues (modifying the totality of how content provider
> pages appear to users), and in the case of systems like Phorm, a
> range of stunningly obvious privacy concerns.
>
> However, the new revelations of security holes introduced by
> ISP-operated DNS diversion systems brings the issue of liability
> front and center.
>
> To the extent that ISPs actually inject (for example) new ads into
> the content streams being sent to users by Web services, it would
> appear that ISPs are taking on an array of liability risks that
> might be associated with those injections.
>
> Possible vectors of risk include security problems associated with
> the injected code, phishing or fraud issues associated with any
> advertisers who are the subject of injected materials, and so on.
> Content providers who might be falsely blamed by users for any such
> problems (which were actually caused by ISP injections) would seem
> to be another potential litigation source.
>
> I've had trouble finding a valid analogy for such injections in
> other contexts.  It clearly doesn't happen with physical mail or
> conventional telephone service.  In television and radio
> broadcasting, the relationships between network and local
> programming, network and locally-inserted commercials, and other
> involved parties, are clearly laid out contractually in virtually
> all cases.
>
> It seems that the ISP push to monitor and/or modify the content of
> user data streams is solely predicated on "we don't think it's
> illegal, so we'll keep pushing the envelope" logic.
>
> Question: Why should users' Internet communications be subject to
> related default conditions that would be viewed as inappropriate in
> most other communications environments?  Can this be reasonably
> justified?  Is it likely to become the subject of civil or criminal
> actions, legislative changes, etc.?
>
> --Lauren--
> NNSquad Moderator