W3C home > Mailing lists > Public > public-tracking@w3.org > March 2012

Re: Proportionate Response for Fraud Prevention and Security (ISSUE-24)

From: Roy T. Fielding <fielding@gbiv.com>
Date: Thu, 15 Mar 2012 12:37:13 -0700
Cc: Tracking Protection Working Group WG <public-tracking@w3.org>
Message-Id: <C43B0934-C582-4753-B73D-AC50B19CC3AF@gbiv.com>
To: Rigo Wenning <rigo@w3.org>
On Mar 15, 2012, at 2:19 AM, Rigo Wenning wrote:

> Roy, Jonathan, Shane, 
> On Wednesday 14 March 2012 11:39:55 Shane Wiley wrote:
>> Please understand these activities are to PROTECT users and businesses alike
>> (depends on the attack).  I'm hopeful we don't purposely create real risk
>> of harm to users in our attempts to "lock down" the DNT standard.
> Security vs Privacy is a big classic in data protection. Our forefathers of 
> data protection in the seventies said that good data protection is requiring 
> more secure systems to protect also against abuse of personal information. So 
> they tried to harmonize security and data protection. 
> On the one hand, I have a lot of sympathy with Roy warning us to open that can 
> of worms. I would be very reluctant to include security-related provisions 
> into the two Specifications. On the other hand, I also have a lot of sympathy 
> for the suggestion to use the present expertise to have some privacy 
> suggestions for the fraud-fighters in the Web's payment channel. 
> Because PROTECT is relative. I'm pretty sure that Assad claims to PROTECT 
> Syria. So only saying "protect" as a use limitation doesn't save our live 
> here.  A best practices document on fraud protection for ad companies would be 
> cool. This could determine unnecessary data collection and identify doubtful 
> sharing practices that would allow to abuse the data collected for fraud 
> protection. In one word, make fraud protection for the web smarter to some 
> extend, privacy wise.. And I think that in a second generation, we could have 
> a framework where a service agrees to back down a bit because the users have 
> decided (via DNT) not to be as highly secured because they favor privacy in a 
> given context.

That's all true, Rigo, but it has nothing to do with DNT.  DNT does not exist
to solve all privacy problems.  User preferences cannot solve security problems.
There is a continual tug of war between security and privacy, yes, but it is
not our war to resolve.  As I said, I strongly encourage regulators to take this
on directly, if they have not done so already, since it is not a matter that
user preferences/consent can resolve.  Fraud control is not about consent.

Data retention for the sake of fraud control should be limited regardless of
the DNT signal because it should be assumed that the user has not consented
(they are usually not given a choice).  I don't know how they should be limited.
I am certain that this working group is incapable of reaching consensus on
how fraud control must be limited, since it depends on the nature of the
fraud, the nature of what is being protected, and the nature of the
organization doing the protection.

In short, we have neither the time, nor the expertise, nor the authority to
address this problem in general, other than to say that:
there exists an exemption for fraud control and data collection/retention/use
under that exemption must be limited to what is necessary for that fraud control.

Received on Thursday, 15 March 2012 19:37:41 UTC

This archive was generated by hypermail 2.3.1 : Friday, 3 November 2017 21:44:46 UTC