> We must also remember that private data can be processed to deliver the
> service. You need to know my address to deliver the parcel, the mobile
> phone company must know my  ID and my location 0cell tower attachment- to
> connect the call etc etc.


Note that much of the research in privacy-enhancing technologies is about 
reducing the information that needs to be known to provide the service. For 
example, there are proposals for anonymous delivery of physical goods 
(https://www.cs.columbia.edu/~smb/papers/APOD_PETS09.pdf). Other examples in 
which the required information is reduced beyond what it would seem intuitive 
include ZK protocols, PIR, etc. 



On 01 Mar 2013, at 16:16:5, Bryan McLaughlin (brmclaug) wrote:

> 
> 
> On 01/03/2013 13:19, "Claudia Diaz" <[email protected]> wrote:
> 
>> 
>> On 01 Mar 2013, at 01:22:12, SM wrote:
>>> At 15:56 26-02-2013, Eric Burger wrote:
>>>> So what if we just said Security Considerations must include what some
>>>> people call privacy considerations? If we cannot agree on a concise
>>>> definition of security vs. privacy, what is the typical draft author
>>>> going to do?
>>> 
>>> Security Considerations can be stuff like "MUST implement TLS".
>>> Privacy considerations would be about the decision of an individual.
>> 
>> 
>> 
>> Framing privacy as "decision-making" is useful when dealing with privacy
>> settings/policies and user interfaces -- this is actually a usual
>> definition of privacy in HCI (see, eg, the work from Cranor or Acquisti
>> on feedback and awareness, privacy nudges, etc.).
>> 
>> I would however think that it is too restrictive (and potentially
>> misleading) when addressing privacy considerations more generally, and
>> particularly in protocols that the individual does not necessarily
>> understand in detail.
>> 
>> BMC>
>> 
>> I would offer that the decision making of the consumer is later in the
>> privacy process especially in the IETF context. Our design process is
>> orthogonal to the privacy choice of the consumer/user upon service
>> signup/purchase.
>> 
>> We must also remember that private data can be processed to deliver the
>> service. You need to know my address to deliver the parcel, the mobile
>> phone company must know my  ID and my location 0cell tower attachment- to
>> connect the call etc etc.
>> 
>> I hallucinate the following process:
>> 
>> We define the data to be collected for a protocol/service/architecture.
>> 
>> We then analyse the resulting list of data and determine if it 'may' be
>> affected by privacy concerns.  - This is where we need to make life easier
>> for ourselves by having  access to globals relevant guidance and keeping
>> it up to date.
>> 
>> We then determine if the context of its use is appropriate within the
>> privacy context.
>> 
>> In the above location example the mobile phone company 'may' not 'need'a
>> more granular location to deliver the service as sold. Particularly if
>> that greater accuracy is to be used to offer value added services or
>> insight into the movement of the subscriber.
>> They may need a more granular location to offer greater level of original
>> service such as informed offload etc.
>> 
>> We can then use the process above to design the architecture to allow the
>> greater level of location processing but also recognise that the service
>> may need to be offered without than granularity.
>> 
> 
> 
>> 
> 


--------------------------------------------------------------------------
Prof. Dr. Ir. Claudia Diaz

Katholieke Universiteit Leuven                       
Dept. Electrical Engineering-ESAT / COSIC
Kasteelpark Arenberg 10, B-3001 Leuven, BELGIUM
tel: +32 16 32 18 14    fax: +32 16 32 19 69

email: claudia.diaz (at) esat.kuleuven.be
web: http://homes.esat.kuleuven.be/~cdiaz/
--------------------------------------------------------------------------

_______________________________________________
ietf-privacy mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ietf-privacy

Reply via email to