Re: [v6ops] Google Alert - IPv6

[Mark Andrews <[e-mail]>]

In message <[e-mail]>, JORDI PALET M
ARTINEZ writes:
> Totally agree, however there is a subtle situation here, when we
> associate this to the origin of the thread  CGN
>
> If an unlawful action has been done with that phone, the phone owner will
> be responsible in front of the law to identify the person who has go the
> phone. Otherwise, the courts will judge you as the author of the unlawful
> action.

Please cite the relevent law.  Phone are hacked all the time.  Phones
are shared all the time.  There is only ever a strong correlation
never absolute proof.  For proof you need other data.

> Same as if you are the owner of a car that has an accident and dont stop,
> unless you identify in and undoubtable way a third person driving the
> car, it will be your problem.

Cars get stolen all the time and sometimes the first you know about
is a knock on the door.  The owner of the vehicle is just the first
step in the investigation.  Things like a speeding fine may get
lumbered on you but failure to stop requires more that you are the
owner of the vechicle to be proven.

> So, and this is what I was trying to point to the EU article 29 working
> party many years ago, even if you identify an IP address, it is almost
> impossible to say this is personal data, you need to correlate that with
> many many many other data, and even do, unless you have a live record of
> he/she being in front of the keyboard at that specific time.
>
> However, the implications of CGN for the police is that, without a CGN,
> the owner for the phone that done that unlawful act (in EU you need to
> provide a legal ID to have a phone line) using that IP address, the
> identification is direct, no further investigation is needed and is the
> owner the responsible to identify a third party if that's the case. With
> CGN, if the IP address is shared and there are no source ports records,
> the investigation brings the police to a number of people, may be only
> 16, may be hundreds, which is a big issue.

Only if you are looking at single events.  Multiple events will
probably whittle the set down to a subscriber.  That doesn't
necessarially identify a individual.

> Regards,
> Jordi


-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: [phone]                 INTERNET: [e-mail]