My thinking, when I first proposed what are now called Warrant Canaries, was to rely on the U.S. constitutional protection against ex post facto. In the same way that you can never know what laws might someday be passed which could criminalize current, lawful, behavior or speech, a service can never know in advance if they might someday receive a warrant or NSL. Can someone be compelled to refrain from free speech because someday that speech might be prohibited? If more than one person is engaged in the creation or publication of a Warrant Canary could it be viewed as a conspiracy? Can someone be compelled to speak falsities under court order?

Until there are prosecutions which reach high enough in the Federal courts we may not know.

Steve

On Wed, Apr 1, 2015 at 8:32 AM, Stephen D. Williams <sdw@lig.net> wrote:
On 4/1/15 5:26 AM, Andrew wrote:
Alfie John:

On Mon, Mar 30, 2015, at 10:01 AM, Steven Schear wrote:

Looks like Australia has banned use of my idea.

http://boingboing.net/2015/03/26/australia-outlaws-warrant-cana.html

If its true that a man's status can be measured by his enemies.. then
I've taken a position at the top of the cypherpunks heap :)

How about the reverse? As the point of canaries is to let people know a
warrant is in place while thinking that you're not breaking any laws by
telling them (good luck), hypothetically why not just be up front and
tell people that a warrant is in place via a tor and a hidden service
(let's call it WarrantWatch). Each post is a message from an admin of a
website saying that a warrant is in place, with the message being signed
via the website's TLS private key for verification.

Alfie


So, you're suggesting that instead of going into a legal 'gray area',
website operators should simply obviously violate the law and then
publish a non-repudiable cryptographic proof of their lawbreaking.

Am I missing something here?  Is the idea to get everyone flouting the
law and thereby render it ineffectual, or is it just April Fools?

Andrew

The point of a warrant canary is to communicate in a legal way something that is illegal for you to communicate directly.  That involves something like setting laws against each other in some prioritized way or otherwise splitting hairs so that no one can technically be prosecuted.  It is a bit of a legal arms race.

You have to find something you cannot be compelled to do or not do.  A dead man's switch of some kind where not doing something is ambiguous and unassailable.  Can you be required by law to go to Starbucks every morning?  To report or not report on Facebook something factual, like you ate Wheaties that morning?  You have a headache?  Something that is protected by a higher priority law, although if the First Amendment is trumped by these laws that's going to be difficult.  Perhaps a group of people probabilistically do or not do something, but any evidence of collusion would be a problem so they would have to act independently.  Some mechanism that relies on someone's thoughts might be sacrosanct.  Visual dwell, polygraph (not that they work at all), etc.

Perhaps every day someone proposes that the warrant canary has triggered, and every day but one someone objects.

Is there or is there not a way to do this legally?  That seems like a gray area, if the First Amendment is trumped, along with other legal protections that could apply.
If there is no way to do this legally, what are the ways to dilute the situation as to be effectively legal, i.e. prosecution would be unlikely or ineffective?
Since the First Amendment is strained here, reasonable people could conclude that the conflicting law is unconstitutional.  That could lead to a sense of responsibility to do the right thing.

sdw