[gnso-rds-pdp-wg] Facebook loses Belgian court case over consent and tracking

Tue Feb 20 18:27:35 UTC 2018

To try to bridge this gap a little bit... and hopefully clarify my point to
Steve.

Shutting down public access to A records degrades the network in a very
immediate, obvious way. Therefore, even if they are PII, they still can be
published.

While WHOIS details are materially different, and shutting down WHOIS
entirely probably won't have an immediate effect, there is an entire
ecosystem of anti-spam, anti-ddos, anti-hack measures in place that allow
Internet users to successfully access their mailboxes, communicate with
remote IP addresses, and so on, without the inundation of spam and DDOS
that would make the entire network unusable if there weren't active efforts
to stop this stuff.

The Internet is a band-aid on top of a band-aid on top of a band-aid. This
WHOIS band-aid turns out to be holding a lot of things together, even
though it was never intended to be originally.

There are uncountable public incidents where abuse caused widespread
Internet shutdown, and was only remediated by active efforts on the part of
network/server operators and anti-abuse professionals. And we are arguing
that WHOIS has often been a critical, and irreplaceable component of many
of those incident responses. We care less about the format of the data, but
we care very deeply that we should still be able to process this data to
respond to future incidents at the same level of effectiveness.

We also argue that our roles cannot be replaced by police that get warrants
and court orders for WHOIS data, for so many reasons, not the least of
which is that no one wants NOCs, SOCs, etc, staffed by cops!

Therefore quite a lot of us are arguing that even though the effect is not
immediate and obvious, the loss of WHOIS will cause secondary effects that
end up degrading the Internet, just like the loss of A records would,
albeit slower and more indirectly. Shut down WHOIS, and you might not
reliably get your A records, your e-mail, or your favorite website.

I think if you spoke to people at large networks, large mailbox operators,
etc, you would find a lot of agreement on the critical importance that
WHOIS has accidentally gained.

On Tue, Feb 20, 2018 at 1:12 PM, John Bambenek via gnso-rds-pdp-wg <
gnso-rds-pdp-wg at icann.org> wrote:

> Which brings me back full circle to my point. An A records exists because
> a registrant put it there. If the system was slightly modified to provide a
> free privacy option, then whois data is there because the registrant put it
> there. And thus the problem is solved.
>
> --
> John Bambenek
>
> > On Feb 20, 2018, at 12:07, Tapani Tarvainen <ncsg at tapani.tarvainen.info>
> wrote:
> >
> > Yes. Even though IP addresses &c can be personally identifiable
> > information, that doesn't mean they can't ever be published. It does
> > mean GDPR applies, but it's clear GDPR would allow DNS records to be
> > published just as they've always been.
> >
> > Seriously. GDPR is not insane.
> >
> > Tapani
> >
> >> On Tue, Feb 20, 2018 at 12:31:57PM -0500, Steve Crocker (
> steve at shinkuro.com) wrote:
> >>
> >> John,
> >>
> >> I think you're making the implicit assumption that access to name server
> >> (NS), address (A and AAAA) and related MX and DS records should be gated
> >> simply because someone can claim there might be personally identifiable
> >> information associated with these records.  This is a very large
> assumption
> >> with very large consequences.
> >>
> >> The DNS was designed to provide unfettered access to these records.  The
> >> implication, therefore, is that anyone who publishes these records
> >> necessarily expects these records to be publicly available.
> >>
> >> If you think there's a need for a system that makes the address
> information
> >> about a site accessible to only a selected set of people, design and
> build
> >> a system that provides that functionality.  The Domain Name System,
> >> however, is not designed and not built that way.  Anyone who publishes
> >> information in the DNS has necessarily chosen to make that information
> >> public.  That's the end of the privacy issue with respect to the Domain
> >> Name System.
> >>
> >> Discussion about how much information about the registrant is a separate
> >> matter, of course.
> >>
> >> Steve
> >>
> >>
> >> On Tue, Feb 20, 2018 at 12:17 PM, John Bambenek <
> jcb at bambenekconsulting.com>
> >> wrote:
> >>
> >>> We have no idea how to determine nationality, so we just assume GDPR
> >>> applies.
> >>>
> >>> We have no idea how to determine natural vs legal person, so we assume
> >>> natural person.
> >>>
> >>> We assume the user is too stupid to use a role-based email address or
> any
> >>> other mitigations, so that’s not an option.
> >>>
> >>> We assume the user is too stupid to know why to do with voluntary
> fields,
> >>> so putting data in whois is too risky even in an opt-in scenario.
> >>>
> >>> But now we are talking about acceptable levels of risk with
> nameservers?
> >>>
> >>> How are we going to control to make sure only the types of data
> processing
> >>> on this sensitive information is limited to what is authorized? What
> if I
> >>> don’t want you to have my nameservers? What are you registrars going
> to do
> >>> to make that possible?
> >>>
> >>> --
> >>> John Bambenek
> >>>
> >>> On Feb 20, 2018, at 11:11, Ayden Férdeline <icann at ferdeline.com>
> wrote:
> >>>
> >>> Domain names and name servers can comprise personal information.
> However,
> >>> this does not mean we cannot use them. It just means we need to
> complete a
> >>> privacy impact assessment and understand the risks involved. And I
> suspect
> >>> the risks to a name server or domain name itself being public are
> >>> incredibly low. The risk profile is nowhere near that of WHOIS or RDS
> being
> >>> open for all to see, filled with sensitive data like addresses and
> phone
> >>> numbers.
> >>>
> >>> Ayden
> >>>
> >>>
> >>> -------- Original Message --------
> >>> On 20 February 2018 5:55 PM, John Bambenek via gnso-rds-pdp-wg <
> >>> gnso-rds-pdp-wg at icann.org> wrote:
> >>>
> >>> Domain names, hostnames, and IP addresses in so far as they are
> personally
> >>> identifiable are PII. Courts have ruled on IP addresses already and
> DPAs
> >>> have said much the same.
> >>>
> >>> So the same logic on why we can’t have a system that lets people
> advertise
> >>> who owns the domain is the same argument why DNS must be gated.
> >>>
> >>> Has any registrar done a PIA on publishing my nameservers? How do I
> >>> control who gets that information? How do we enforce its for authorized
> >>> purposes only?
> >>>
> >>> --
> >>> John Bambenek
> >>>
> >>> On Feb 20, 2018, at 10:50, Steve Crocker <steve at shinkuro.com> wrote:
> >>>
> >>> I'm puzzled by the reference to name servers and A records.  These are
> >>> necessarily public else the domain name system won't function.  Is
> there
> >>> confusion or misunderstanding about the role of these records?
> >>>
> >>> Steve
> >>>
> >>>
> >>>> On Tue, Feb 20, 2018 at 11:47 AM, allison nixon <elsakoo at gmail.com>
> wrote:
> >>>>
> >>>> 1,000,000% agreed. Registrars cannot eliminate all their risk by
> masking
> >>>> WHOIS into oblivion. The DPAs can still ask why they are exposing A
> >>>> records, nameservers, etc, to anyone who asks for them, without valid
> >>>> reasons or authentication. Why do they expose zone files, etc. The
> DPAs can
> >>>> ask why customer support can sometimes so easily be social engineered
> into
> >>>> handing over accounts to account takeover scammers.
> >>>>
> >>>> Since most registrars are also hosting providers/mail providers, would
> >>>> criminals storing stolen PII on your servers be a GDPR issue? After
> all,
> >>>> the ultimate owner of the server is also considered a "processor",
> which
> >>>> has interesting implications if one's customers include phishers, or
> sell
> >>>> stolen credit cards, and one's already been notified. I have even seen
> >>>> miscreants putting doxes in TXT records.
> >>>>
> >>>> I already know of quite a few incidents where people would have had
> >>>> standing to file a GDPR complaint against registrars/hosters,
> unrelated to
> >>>> WHOIS.
> >>>>
> >>>> Eventually the issue is going to impact the core business model of
> >>>> registrars. This isn't going to stop at WHOIS. An open dialog with
> the DPAs
> >>>> at an early stage is of utmost importance for all parties involved
> here.
> >>>>
> >>>>
> >>>> On Mon, Feb 19, 2018 at 10:16 AM, Sam Lanfranco <sam at lanfranco.net>
> >>>> wrote:
> >>>>
> >>>>> Benny,
> >>>>>
> >>>>> This is why I support multi-venue multi-stakholder dialogue with the
> >>>>> DPA's so that they are appraised of the issues on all sides of the
> data
> >>>>> protection issue. They are then more likely to act in a judicious
> manner,
> >>>>> and less like an attack dog. Watch the new movie "*The Post*" where
> >>>>> when *Washington Post* owner Katharine Graham decided to publish the
> >>>>> Vietnam War Pentagon Papers, with the downside risk that she could be
> >>>>> jailed for treason. The court ruled in favor of freedom of the
> press. It is
> >>>>> not what the DPA can do, but what they are likely to do, and
> dialogue goes
> >>>>> a long way to mitigating risk and shaping appropriate positions and
> >>>>> behavior (with integrity) on all sides.
> >>>>>
> >>>>> Sam L.
> >>>>>
> >>>>> On 2/19/2018 10:02 AM, benny at nordreg.se wrote:
> >>>>>
> >>>>> <ironi on> Now I am relieved, we as registrars will not be subject
> for
> >>>>> anything… </ironi off>
> >>>>>
> >>>>> None of us know where and what they will prioritise,* remember that
> it
> >>>>> only take 1 complaint to a DPA to get the snowball moving.* [emphasis
> >>>>> added] I am sure your statement have noe value then.
> >>>>>
> >>>>> --
> >>>>> Med vänliga hälsningar / Kind Regards / Med vennlig hilsen
> >>>>>
> >>>>>
> >>>>> Benny Samuelsen
> >>>>> Registry Manager - Domainexpert
> >>>>>
> >>>>> Nordreg AB - ICANN accredited registrar
> >>>>> IANA-ID: 638
> >>>>> Phone: +46.42197000 <+46%2042%2019%2070%2000>
> >>>>> Direct: +47.32260201 <+47%2032%2026%2002%2001>
> >>>>> Mobile: +47.40410200 <+47%20404%2010%20200>
> >>>>>
> >>>>> On 19 Feb 2018, at 15:29, Sam Lanfranco <sam at lanfranco.net> wrote:
> >>>>>
> >>>>> Hi Tim,
> >>>>>
> >>>>> No, completely to the contrary. My point with that dollars reference
> was
> >>>>> that in some cases litigation is the preferred business response,
> rather
> >>>>> than compliance and paying fines. Also, the big revenues in mining
> big data
> >>>>> are outside the DNS sphere, and outside the abuses and "bad things"
> that
> >>>>> websites do to people. The big EU fines are more likely to hit
> social media
> >>>>> than Registrars, although they are risks there as well. The
> revenues, and
> >>>>> privacy violations, will come from profiling users by mining big
> data for
> >>>>> scraps of personal date to individualize target marketing.
> >>>>>
> >>>>> *As a brief aside:* This goes well beyond the remit of ICANN and is
> >>>>> actually worse than just being inundated by adverts base on personal
> online
> >>>>> behavior. Artificial Intelligence mining apps are increasingly
> customizing
> >>>>> the "news" one gets from news feeds, to help "glue the eyeballs" to
> the
> >>>>> adverts, creating a news silo of one.  (That is amusing for me since
> I
> >>>>> virtually live in two towns in two countries). Even more worrisome
> is the
> >>>>> growing practice for A.I. companies where A.I. "writes" the news
> releases,
> >>>>> now mainly in sports and finance, for thousands of print and online
> news
> >>>>> outlets. I know all of this is outside the ICANN remit so I will stop
> >>>>> there.
> >>>>>
> >>>>> Sam L.
> >>>>>
> >>>>> On 2/18/2018 5:43 PM, Chen, Tim wrote:
> >>>>>
> >>>>> Hi Sam,
> >>>>>
> >>>>> When you say these are hundred million dollar issues for "the
> >>>>> companies",which companies are you talking about?  Large Registrars?
> >>>>>
> >>>>> I hope you are not comparing cybersecurity professionals and the good
> >>>>> work they are trying to enable, to a completely separate privacy
> issue
> >>>>> around data used for ad tracking or behavior tracking across
> websites.  If
> >>>>> I spent my days trying to protect people on the internet from bad
> things, I
> >>>>> would certainly not appreciate any allusion that I was engaged on
> the whois
> >>>>> data issue 'for the money'.
> >>>>>
> >>>>> Tim
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> gnso-rds-pdp-wg mailing list
> >>>>> gnso-rds-pdp-wg at icann.org
> >>>>> https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg
> >>>>>
> >>>>>
> >>>>>
> >>>>> --
> >>>>> ------------------------------------------------
> >>>>> "It is a disgrace to be rich and honoured
> >>>>> in an unjust state" -Confucius
> >>>>> 邦有道，贫且贱焉，耻也。邦无道，富且贵焉，耻也
> >>>>> ------------------------------------------------
> >>>>> Visiting Prof, Xi'an Jaiotong-Liverpool Univ, Suzhou, China
> >>>>> Dr Sam Lanfranco (Prof Emeritus & Senior Scholar)
> >>>>> Econ, York U., Toronto, Ontario, CANADA - M3J 1P3
> >>>>> email: sam at lanfranco.net   Skype: slanfranco
> >>>>> blog:  https://samlanfranco.blogspot.com
> >>>>> Phone: +1 613-476-0429 <(613)%20476-0429> cell: +1 416-816-2852
> <(416)%20816-2852>
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> gnso-rds-pdp-wg mailing list
> >>>>> gnso-rds-pdp-wg at icann.org
> >>>>> https://mm.icann.org/mailman/listinfo/gnso-rds-pdp-wg
> >>>>>
> >>>>
> >>>>
> >>>>
> >>>> --
> >>>> _________________________________
> >>>> Note to self: Pillage BEFORE burning.
> > _______________________________________________
> > 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
>

-- 
_________________________________
Note to self: Pillage BEFORE burning.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/gnso-rds-pdp-wg/attachments/20180220/90246f3f/attachment-0001.html>