[gnso-rds-pdp-wg] Now open: 18 January Poll on Purpose

Wed Jan 25 21:45:23 UTC 2017

I don't know if it will be helpful to anyone else, but I found it helpful to me to go back review why we are talking about purpose.

******** Charter question 1 is titled Users/Purposes and says: "Who should have access to gTLD registration data and why?" To answer the 'why' part of the question leads us to a discussion of purpose, albeit purpose for access, not for collection.  I would argue though that if we agree on a purpose for access, then the data will have to be collected.

******** The sub-questions under question 1 came directly from the Issues Report (see pages 39-41 & 70) and were included in the Mind Map of our work. Note what they say (bolding done by me):

o   Should gTLD registration data be accessible for any purpose or only for specific purposes?

o   For what specific purposes should gTLD registration data be collected, maintained, and made accessible?

o   What should the over-arching purpose be of collecting, maintaining, and providing access to gTLD registration data?

o   What are the guiding principles that should be used to determine permissible users and purposes, today and in the future?

Our charter clearly directs us to discuss purpose and it clearly talks about collection.  Should we have talked about access first?  Maybe.  But we will get there fairly soon.  Should we talk about purposes for access  instead of purpose of the RDS?  Probably.

Chuck

From: James Galvin [mailto:jgalvin at afilias.info]
Sent: Wednesday, January 25, 2017 8:52 AM
To: Gomes, Chuck <cgomes at verisign.com>
Cc: RDS PDP WG <gnso-rds-pdp-wg at icann.org>
Subject: [EXTERNAL] Re: [gnso-rds-pdp-wg] Now open: 18 January Poll on Purpose

Chuck,

While what you say is reasonable it also shows that you, as well as myself and the rest of us, are stuck blurring the lines between collection, storage, and display. I want to try a different approach.

You start off saying "collection for use in the RDS". Then you say "collected for the RDS". Then you say "needed in the RDS".

Just to confuse things I'm going to assert that all those phrases are wrong.

The RDS is not the purpose here. It is a solution for the purpose. It is a system, a principle, or a concept. The details of its implementation are going to be interesting for a number of reasons but let's stick to the high-level for a moment.

The purpose of the RDS is to support the management or control of the lifecycle of a domain name. For me, this is self-evident and really should not be subject to much discussion. If you disagree with this then we are creating a more typical cash-based retail system such as might exist for selling chewing gum, e.g., no registries and no registrars. We know this won't work because there are certain properties we need from our domain names, not the least of which is global uniqueness. Thus, there needs to exist an RDS that we can all share so we can all do our part to preserve an appropriate set of properties of domain names.

I believe the question we are considering is if there are any other purposes of the RDS for which ICANN (either the whole or a subset of the royal ICANN including the corporation, contracted parties, interested parties, and random Internet users) is a required participant and thus use of the RDS should be required for those purposes and thus there are requirements on the data that should be present in the RDS. The questions of access and collection are an entirely different discussion.

While I agree that most of the "purposes of thin data" suggested on the call yesterday are reasonable and sensible, perhaps even important, I am not yet convinced they rise to the level of requiring ICANN (the royal ICANN) to be an active participant and thus are not an appropriate purpose of the RDS.

This is not about collection, storage, or display, or its about all of them. Why is ICANN and the RDS essential to the suggested purposes?

The fact that data collected on behalf of managing the lifecycle of a domain name just happens, coincidentally, to be useful for other purposes, is just a happy circumstance. We can separately have discussions about how the data that is being collected could be used by others. I'm troubled by the fact that it is driving the current discussion.

Jim

On 24 Jan 2017, at 19:57, Gomes, Chuck wrote:

Regarding our focus on collection and also on uses versus purposes, I would like to share some thoughts that I shared privately with a WG member a few days ago.

When I say 'collection of data', I mean collection for use in the RDS regardless of how it might be displayed or accessed.  In our leadership call yesterday, Michele pointed out that thin data is generated, not collected.  From the point of view of registries and registrars, I believe that is correct.  But as others have pointed out, our task is not to decide what registries and registrars collect but rather what would be collected for the RDS.  Of course, if we develop policies that require that a thin data element is needed in the RDS, it would need to be collected from registries and/or registrars, and it would then follow that registries and/or registrars would have to generate it.

With that explanation of my use of the word 'collect', let me also make a few other points:

*        We initially narrowed our focus on 'collection only' to keep our focus on a more manageable scope; right or wrong, we decided that focusing on collection and display and access all at once would make our task more difficult.

*        One problem with focusing on 'collection only' is that it causes us to reach conclusions that may later be changed and I am suspecting that that is part of the problem.  For example, when we conclude that we have reached rough  consensus that research is a legitimate purpose for collecting thin data, that conclusion could later change when we talk about display of certain thin data elements or access to some thin data elements.  It might be better if we called our conclusions 'interim conclusions' because they may change after deeper deliberation.

To take this one step further, in our further deliberations on data elements beyond collection we could find that there are no legitimate uses for particular data elements.  In such cases I think that we would then delete any interim purposes for which we reached rough consensus for collection of them.  In other words, if there is no legitimate use for the data, whether public or gated access, it seems to me that there would be no need to collect it.

Does any of this make sense?

Chuck

From: gnso-rds-pdp-wg-bounces at icann.org<mailto:gnso-rds-pdp-wg-bounces at icann.org> [mailto:gnso-rds-pdp-wg-bounces at icann.org] On Behalf Of James Galvin
Sent: Tuesday, January 24, 2017 4:20 PM
To: Greg Aaron <gca at icginc.com<mailto:gca at icginc.com>>
Cc: RDS PDP WG <gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>>
Subject: [EXTERNAL] Re: [gnso-rds-pdp-wg] Now open: 18 January Poll on Purpose

I'm still struggling to understand your answer to my question, "How are the rights of legitimate users infringed by requiring them to authenticate in order to get access to data?"

Scott has reframed my question by observing that one consideration for this group is that there may be data that is available anonymously and data that requires authentication of the requestor. Either that consideration implicitly infers that rights are not infringed or we still have a question to answer.

Taking a step back, it's probably too soon to focus too much on this issue. We're really talking about data collection at this time so the parameters of display or access are relevant but don't need to be solved just yet.

I agree there are a number of interesting if not important use cases. The question I am pressing is why these use cases justify collection of the data? If we can't clearly answer this question then I will continue to object to collecting the data for the purpose being discussed.

I'm wondering if you're asserting that legacy uses establish legitimate usage and since that has been anonymous to date it is an infringement of a "right" to take that away.

Jim

On 24 Jan 2017, at 14:05, Greg Aaron wrote:

I made a typo below.  It should read: "For example law enforcement investigators do not want to reveal what they are looking into, for obvious reasons.  I also assume that they do not want to violate terms of service or lie about who they are or what they are doing."

--Greg

From: gnso-rds-pdp-wg-bounces at icann.org<mailto:gnso-rds-pdp-wg-bounces at icann.org> [mailto:gnso-rds-pdp-wg-bounces at icann.org] On Behalf Of Greg Aaron
Sent: Tuesday, January 24, 2017 1:53 PM
To: James Galvin <jgalvin at afilias.info<mailto:jgalvin at afilias.info>>; RDS PDP WG <gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>>
Subject: Re: [gnso-rds-pdp-wg] Now open: 18 January Poll on Purpose

Dear Jim:

If legitimate users must identify themselves before looking up a domain name, and are required to state their purpose for making that query, that is a significant collection of data with huge privacy and security implications.  It means that registrars and registry operators would collect information about what specific people are searching for, and why.   Users who have perfectly legitimate uses are not currently required to give up their identities and use cases - can such a change be justified?

As a practical matter, a change would makes people jump through hoops unnecessarily.  If we're talking about thin data, that data is not sensitive or personally identifiable.  Thus there's no reason for people who want to access it to declare their identities and use cases.

Currently our RDS system (WHOIS) is a public query/response system.  You're pointing to turning RDS into a credential-driven system.  That poses enormous consequences for privacy, security, and cost.  A lot of people commented about those things in response to the EWG.  See SAC061 or example.

There are use cases that argue in favor of anonymous access.  For example law enforcement investigators do not want to reveal what they are looking into, for obvious reasons.  I also assume that they do they want to violate terms of service or lie about who they are or what they are doing.

The setup we currently have -- a query/response system that does not require credentials or permissions -- avoids the above problems, among others.

All best,

--Greg

From: James Galvin [mailto:jgalvin at afilias.info]
Sent: Tuesday, January 24, 2017 12:55 PM
To: RDS PDP WG <gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>>
Cc: Greg Aaron <gca at icginc.com<mailto:gca at icginc.com>>
Subject: Re: [gnso-rds-pdp-wg] Now open: 18 January Poll on Purpose

I have a comment and a question about Greg's suggestion/conclusion.

First, while I appreciate the documented history and recollection of precedent, we are frequently reminded that we are starting with a clean slate. Thus the fact that "all use is allowed except when a use is specifically prohibited" currently exists in contracts is not binding for us.

Second, could you say more about how the rights of legitimate users are infringed by having to identify themselves before getting access to data? In my experience it is ordinary process to identify yourself for access (and sometimes authenticate yourself) unless the circumstances are known to be anonymous or public. This group has to decide if the circumstances justify anonymous or public access, and what that means.

Thanks,

Jim

On 23 Jan 2017, at 12:45, Greg Aaron wrote:

   The question is not "Is ICANN a law enforcement body?''  (Clearly it is not.)  The question is whether ICANN can require that data be collected and published in order to facilitate various legitimate goals.  The answer to that question is clearly "yes."

   ICANN's Bylaws describe ICANN's responsibilities and their scope - especially see Article 1, section 1, of ICANN's Bylaws, entitled "Mission, Commitments and Core Values."  (https://www.icann.org/resources/pages/governance/bylaws-en/#article1 )   Among other things, "ICANN's scope is to coordinate the development and implementation of policies: For which uniform or coordinated resolution is reasonably necessary to facilitate the openness, interoperability, resilience, security and/or stability of the DNS including, with respect to gTLD registrars and registries... functional and performance specifications for the provision of registrar services; registrar policies reasonably necessary to implement Consensus Policies relating to a gTLD registry;  resolution of disputes regarding the registration of domain names...  Examples of the above include, without limitation: ...maintenance of and access to accurate and up-to-date information concerning registered names and name servers".

   Years back it was decided that the collection and publication of the data was important for accomplishing some legitimate goals, in keeping with the above principles.  And since the old days there have been additional statements of note.  For example in 2007 the GAC weighed in recognizing a number of specific legitimate uses, including "facilitating inquiries and subsequent steps to conduct trademark clearances and help counter intellectual property infringement" and "contributing to user confidence in the Internet", in keeping with law.  We're reviewing all this now; just saying that there's a lot of precedent, and proposals for chnage need to address precedent.

   Currently there is an approach that's important to mention.  The contracts say that registrars "shall permit use of data it provides in response to queries for any lawful purposes". [Emphases added; and except for "mass unsolicited, commercial messages" i.e. spamming, and some high-volume queries.)   Access is not prohibited or regulated.  All use is allowed except when a use is specifically prohibited.

   The alternative is to enumerate all allowable uses  and to regulate access based on each user's intent to honor those allowed uses.  And that takes the world to a place where a system must gatekeep all users, and parcel out data to them only after the assert or prove they have a legitimate use and that they will employ the data only for that  purpose.  IMHO that infringes upon the rights of legitimate users, and is also a completely unmanageable solution.

   All best,

   --Greg

   From: gnso-rds-pdp-wg-bounces at icann.org<mailto:gnso-rds-pdp-wg-bounces at icann.org> [mailto:gnso-rds-pdp-wg-bounces at icann.org] On Behalf Of Kimpian Peter
   Sent: Monday, January 23, 2017 8:53 AM
   To: Stephanie Perrin <stephanie.perrin at mail.utoronto.ca<mailto:stephanie.perrin at mail.utoronto.ca>>; gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>
   Subject: Re: [gnso-rds-pdp-wg] FW: Now open: 18 January Poll on Purpose

   Dear All,

   Adding to the purpose debate: usually it is common sense and wiedly reckognised that we don't collect personal data just for the sake of it or in bulk saying it will be good for one purpose or another. Usually data controllers have the obligation to say openly in advance this is why I am going to process (ie collect, agregate, transfer, etc.) personal data. Being said that it can not be excluded that those data will be used/accessed for "higher" common good and for the benefit for all by another athorised data controller. For example a telco company can if all conditions met disclose (!) data it previously collected to law enforcement agencies but this does not mean that the Telco compony can collect, process etc data for law enforcement purpose...

   My simple question to start with would be and always was: Is ICANN a law enforcement body? Does ICANN have any power/competence in fighting against crime? And it goes for other purposes as well: Is ICANN an international trademark organisation? Etc...Is the answer given to those questions is shared by all the community of ICANN? In my sense we have to be sure that we answer first those questions before deciding on possible purposes (which does not mean that discloser of data on a case-by case base according to international legal requirements will not be possible after this, but those will be exceptions !!!)

   Best regards,

   Peter

   2017.01.22. 19:20 keltezéssel, Stephanie Perrin írta:

      I love your analogy Shane, it is perfect.  In data protection terms that would be a use.  For a legitimate purpose... sledding.  There might have to be repercussions if you cracked the lid....that might be a data breach:-)

      I hate being a nit picker and calling out this distinction between purpose of collection as opposed to purpose for use and disclosure, but it is extremely important in terms of data protection.  Some laws are more clear than others on the distinction, and you are correct that if we are not careful DP laws will forbid the collection and disclosure of the data.  It is certainly clear that for collection of thin data, there is ample justification for collecting the info based on ICANN's limited mandate.  However adding law enforcement and other similar website related investigative activities to the list of legitimate purposes is in my view opening a barn door.  After a year of discussion we may understand the nuance, that we are talking about thin data, etc etc but when the fruits of our labours are published, it looks like we have all agreed that law enforcement (eg) is a legitimate purpose for collecting registration data.  In my view, it is not.

      cheers Stephanie

      On 2017-01-22 04:03, Shane Kerr wrote:

         Greg,

         If we can say that not all legitimate purposes have to be catered for,
         then I agree with you. :)

         If we say that tracking down the registrar of a domain as part of
         trademark research is a legitimate purpose, that does not mean that we
         have to design the system for this purpose, right?

         To try an analogy: We can recognize that using the plastic top of a
         garbage can as a sled is legitimate, but we don't insist on designing
         lids with sledding in mind.

         Full disclosure: My own take on the "legitimate purpose" discussion
         with regards to "thin data" is that we need *some* purpose for both
         gathering and publishing the information, because otherwise privacy
         laws may prohibit companies from gathering or publishing it. Luckily I
         think that there are so many such purposes that the need for the
         information is indisputable.

         Jumping ahead... as I said in a prior call (sorry for missing ones since
         then), I would prefer that the information is then allowed for any
         purpose, without restriction, because otherwise you have to have not
         only tiresome rules about what is allowed but also the Internet Police
         to enforce those rules, which seems like a step towards Armageddon.

         Given that we're still talking about "thin data", which is basically
         just a pointer to a registrar who has *actual* data, my own
         recommendation is not to stress too much. This stuff is only very, very
         vaguely personally identifiable.

         Cheers,

         --
         Shane

         At 2017-01-21 14:51:29 -0500
         Greg Shatan <gregshatanipc at gmail.com><mailto:gregshatanipc at gmail.com> wrote:

            I have to disagree.  These are legitimate purposes for collection, as well
            as for disclosure.

            Greg

            On Fri, Jan 20, 2017 at 7:02 PM, Stephanie Perrin <
            stephanie.perrin at mail.utoronto.ca<mailto:stephanie.perrin at mail.utoronto.ca>> wrote:

               I filled it out, but I am afraid for most of the purposes I could not
               agree.  We do not *collect *data for many of those purposes.  We disclose
               it to people for those purposes, but the purpose of collecting those data
               elements is not for tax collection, trademark enforcement actions, etc.
               This is the conflation issue I have raised repeatedly.

               Apologies if I did not make that point clear enough on the call.

               Stephanie Perrin

               On 2017-01-20 17:35, Gomes, Chuck wrote:

               Please note that our current poll ends in about 24 hours.  So far only 16
               people have responded.

               Chuck

               *From:* gnso-rds-pdp-wg-bounces at icann.org<mailto:gnso-rds-pdp-wg-bounces at icann.org> [mailto:gnso-rds-pdp-wg-
               bounces at icann.org<mailto:bounces at icann.org> <gnso-rds-pdp-wg-bounces at icann.org><mailto:gnso-rds-pdp-wg-bounces at icann.org>] *On Behalf Of *Lisa
               Phifer
               *Sent:* Wednesday, January 18, 2017 1:50 PM
               *To:* RDS PDP WG <gnso-rds-pdp-wg at icann.org><mailto:gnso-rds-pdp-wg at icann.org> <gnso-rds-pdp-wg at icann.org><mailto:gnso-rds-pdp-wg at icann.org>
               *Subject:* [EXTERNAL] [gnso-rds-pdp-wg] Now open: 18 January Poll on
               Purpose

               Dear all,

               As directed in the 18 January WG call, this week's new Poll on Purpose is
               now open for WG member participation:

               https://www.surveymonkey.com/r/SZX9QJZ

               A PDF of this poll's questions and notes/recordings of the meeting are
               posted on the 18 January meeting page: https://community.icann.org/x/
               EbTDAw

               This poll will close at *COB Saturday 21 January 2017*.

               All WG members are encouraged to participate in this poll to help advance
               deliberation and prepare for next week's meeting.

               Best regards,
               Lisa

               _______________________________________________
               gnso-rds-pdp-wg mailing listgnso-rds-pdp-wg at icann.orghttps://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg<mailto:listgnso-rds-pdp-wg at icann.orghttps://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg>

               _______________________________________________
               gnso-rds-pdp-wg mailing list
               gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>
               https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg

               _______________________________________________
               gnso-rds-pdp-wg mailing list
               gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>
               https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg

      _______________________________________________
      gnso-rds-pdp-wg mailing list
      gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>
      https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg

   _______________________________________________
   gnso-rds-pdp-wg mailing list
   gnso-rds-pdp-wg at icann.org<mailto:gnso-rds-pdp-wg at icann.org>
   https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/gnso-rds-pdp-wg/attachments/20170125/cabb7173/attachment.html>