[gnso-rds-pdp-wg] [EXT] Re: international law enforcement association resolution regarding domain registration data

Tue Apr 25 08:58:08 UTC 2017

Hi Scott, thanks for the reminder of the handle ids.

My suggestion was actually to go beyond the current contact identifier 
and use unique identifiers for more contact fields.
Data anonymisation is very common for a lot of smartphone apps in 
combination with the location of users and used a lot in different 
industries.

Infact, the GDPR and the Eprivacy regulation describe what legal options 
there are when it comes to abuse mitigation. Did we have input from 
Europol on how they deal with the current EU directives? During 
Copenhagen, the person from Interpol explained rather well how they were 
in compliance with privacy laws around the globe and is part of their 
compliance procedure and catching criminals.

Best,

Theo

Hollenbeck, Scott via gnso-rds-pdp-wg schreef op 2017-04-24 05:56 PM:
> Sorry, I was only responding to Theo's question about unique anonymous
> identifiers, and noting that we already have them at our disposal. The
> topics you're describing below are part of a different discussion.
> 
> Scott
> 
> FROM: Paul Keating [mailto:Paul at law.es]
> SENT: Monday, April 24, 2017 11:31 AM
> TO: Hollenbeck, Scott <shollenbeck at verisign.com>;
> 'cdoman at alienvault.com' <cdoman at alienvault.com>;
> 'gnso-rds-pdp-wg at icann.org' <gnso-rds-pdp-wg at icann.org>
> SUBJECT: [EXTERNAL] Re: [EXT] Re: [gnso-rds-pdp-wg] international law
> enforcement association resolution regarding domain registration data
> 
> Scott,
> 
> I understand the context.  However, I am not sure how the RDS gets you
> any further unless it is coupled with a requirement that all
> registrations be verified so that the gated access will result in
> meaningful information.
> 
> Also, I am not sure to what extent it would aid in the accuracy of the
> underlying WHOIS data.  People who want to keep their information
> confidential or who would otherwise provide fake data will certainly
> find a way of avoiding having to provide the data you are seeking.
> Nominet recently (3 years ago :-( ) went through this with the .UK
> requirement of confirmed WHOIS.  They wanted to make sure that
> registrants were "connected" to the UK and that the WHOIS data was
> correct.  However, it is a simple process to avoid, including the use
> of agents and/or dropbox locations and the like.  Corporate registrant
> data remains virtually untraceable (just as real estate holdings do)
> in terms of the Beneficial Owners.  So, while I sympathize with the
> desire to have accurate WHOIS data I remain at a loss as to how this
> can be achieved.  Asking Registrars to employ complex and costly
> investigative or confirmation systems would seem to conflict with the
> economic  model in which registrars receive very little gross margin
> (let alone profit) from domain registrations.
> 
> And, while I understand the argument about privacy when it comes to
> registration data I do not share the opinion that it should all be
> private.  Those desiring to restrict their personal data can easily do
> so through use of agents and privacy services.  That they chose not to
> do so is not an issue that IMHO the rest of the public really needs to
> bear the burden for.
> 
> Disclosure:  I am a lawyer in the domain space, a director of
> DomainTools and an active domain investor.  I use WHOIS in my daily
> practice.
> 
> Paul Keating
> 
> FROM: "Hollenbeck, Scott" <shollenbeck at verisign.com>
> DATE: Monday, April 24, 2017 at 4:54 PM
> TO: "'cdoman at alienvault.com'" <cdoman at alienvault.com>, Paul Keating
> <paul at law.es>, "'gnso-rds-pdp-wg at icann.org'"
> <gnso-rds-pdp-wg at icann.org>
> SUBJECT: RE: [EXT] Re: [gnso-rds-pdp-wg] international law enforcement
> association resolution regarding domain registration data
> 
>> Chris, in the note below I shared an actual example of an existing
>> contact identifier/handle, "C270-LRMS". It's an identifier that's
>> assigned by the registry when the contact is created, and it's
>> guaranteed to be unique within the registry namespace. It can be
>> generated in many different ways (hash values are certainly one
>> possible way) as long as the final value conforms to the syntax
>> described in the EPP specifications.
>> 
>> One thing that makes these identifiers interesting is that they,
>> too, can be searched for in an RDDS. In theory, we could have a
>> system that returned nothing but these abstract identifiers in
>> responses to domain name queries. One would then have to issue
>> additional queries to see the details behind the handle, and if the
>> system were designed in such a way it would be possible to restrict
>> access to this information based on whatever policies exist for
>> access to personally identifiable information.
>> 
>> Scott
>> 
>> FROM: Chris Doman [mailto:cdoman at alienvault.com]
>> SENT: Monday, April 24, 2017 9:38 AM
>> TO: Paul Keating <Paul at law.es>; Hollenbeck, Scott
>> <shollenbeck at verisign.com>; 'gnso-rds-pdp-wg at icann.org'
>> <gnso-rds-pdp-wg at icann.org>
>> SUBJECT: [EXTERNAL] Re: [EXT] Re: [gnso-rds-pdp-wg] international
>> law enforcement association resolution regarding domain registration
>> data
>> 
>> Can you provide an example of a handle please for my understanding?
>> 
>> Do you mean for example a hashed user id from a registrar, or an
>> email address?
>> 
>> -------------------------
>> 
>> FROM:gnso-rds-pdp-wg-bounces at icann.org
>> <gnso-rds-pdp-wg-bounces at icann.org> on behalf of Paul Keating
>> <Paul at law.es>
>> SENT: 24 April 2017 14:31:21
>> TO: Hollenbeck, Scott; 'gnso-rds-pdp-wg at icann.org'
>> SUBJECT: [EXT] Re: [gnso-rds-pdp-wg] international law enforcement
>> association resolution regarding domain registration data
>> 
>> I would like to understand this further and forgive my ignorance.
>> Could
>> you please explain what you mean by the following:
>> 
>> We already have "anonymized unique identifiers" in the form of
>> contact
>> identifiers, sometimes also known as "handles" (but not to be
>> confused
>> with handle system (RFC 3650) identifiers). For example, a WHOIS
>> query for
>> a particular domain to one particular thick RDDS registry service
>> will
>> return a registrant identifier of "C270-LRMS" in addition to the
>> more
>> identifiable information that we're all familiar with. RDAP also
>> supports
>> these identifiers, so they are available for purposes as we see fit
>> to
>> recommend.
>> 
>> Also, although WHOIS information itself is not always helpful, WHOIS
>> combined with other data (including DNS and historical data) can be
>> extremely helpful.  So, I am not at all certain as to what is
>> intended by
>> the message or what the solution proposed is or how it works.
>> 
>> Thank you,
>> 
>> Paul Keating
>> 
>> On 4/24/17, 2:47 PM, "Hollenbeck, Scott via gnso-rds-pdp-wg"
>> <gnso-rds-pdp-wg-bounces at icann.org on behalf of
>> gnso-rds-pdp-wg at icann.org>
>> wrote:
>> 
>>> I must have missed this note when it was original sent back in
>> March.
>>> Chuck asked me to address one of the questions posed below.
>>> 
>>> Scott
>>> 
>>>> -----Original Message-----
>>>> From: Gomes, Chuck
>>>> Sent: Monday, April 24, 2017 7:48 AM
>>>> To: Hollenbeck, Scott <shollenbeck at verisign.com>
>>>> Subject: FW: [EXTERNAL] Re: [gnso-rds-pdp-wg] international law
>>>> enforcement association resolution regarding domain registration
>> data
>>>> 
>>>> FYI Scott.
>>>> 
>>>> Chuck
>>>> 
>>>> -----Original Message-----
>>>> From: theo geurts [mailto:gtheo at xs4all.nl]
>>>> Sent: Thursday, March 02, 2017 4:21 PM
>>>> To: Gomes, Chuck <cgomes at verisign.com>; Greg Aaron
>> <gca at icginc.com>
>>>> Cc: gnso-rds-pdp-wg at icann.org
>>>> Subject: [EXTERNAL] Re: [gnso-rds-pdp-wg] international law
>> enforcement
>>>> association resolution regarding domain registration data
>>>> 
>>>> 
>>>> Thanks, Chuck.
>>>> 
>>>> I think it is important that we as a WG understand that gated
>> access
>>>> could
>>>> be a recommendation. But it does not single out any other
>>>> solutions/recommendations,  but to get to that point, we should
>> keep
>>>> exploring.
>>>> 
>>>> To give this some more color. In 2016 we assisted
>> paloaltonetworks.com
>>>> and
>>>> shadow server taking down the Prince of Persia malware that went
>>>> undetected and roamed the internet for ten years (that's a long
>> time
>>>> folks)
>> http://researchcenter.paloaltonetworks.com/2016/06/unit42-prince-
>> [1]
>>>> of-persia-game-over/
>>>> 
>>>> So the actual WHOIS data was useless in a sense we were dealing
>> with
>>>> stolen identities.
>>>> But we were able to map out the botnet controller network through
>> the
>>>> WHOIS and coordinated with more Registrars to sinkhole the entire
>> lot.
>>>> 
>>>> Again the WHOIS data was useless in this case as it was fake,
>> could have
>>>> passed every syntax or WHOIS cross-validation check.
>>>> 
>>>> So instead of gated access, why not aim for an RDS that used
>> anonymized
>>>> unique identifiers that are available for everyone?
>>> 
>>> We already have "anonymized unique identifiers" in the form of
>> contact
>>> identifiers, sometimes also known as "handles" (but not to be
>> confused
>>> with handle system (RFC 3650) identifiers). For example, a WHOIS
>> query
>>> for a particular domain to one particular thick RDDS registry
>> service
>>> will return a registrant identifier of "C270-LRMS" in addition to
>> the
>>> more identifiable information that we're all familiar with. RDAP
>> also
>>> supports these identifiers, so they are available for purposes as
>> we see
>>> fit to recommend.
>>> 
>>> Scott
>>> _______________________________________________
>>> gnso-rds-pdp-wg mailing list
>>> 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
>> https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg
> 
> 
> Links:
> ------
> [1] http://researchcenter.paloaltonetworks.com/2016/06/unit42-prince-
> _______________________________________________
> gnso-rds-pdp-wg mailing list
> gnso-rds-pdp-wg at icann.org
> https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg