|
||||||||
|
|
|||||||
![]() |
|
|
Thread Tools | Display Modes |
|
#1
|
|
It is starting to look as if BT's involvement in this was greater that
it has been reported so far, possibly to the extent of being the primary development partner. Statis Scleparis, who was BT Retail's Chief Technology Officer between 2006 and 2007 is now the CTO in Phorm. Evidence here - http://www.phorm.com/about/exec_scleparis.php It appears that Scleparis may have been the Retail CTO during last year's covert trials of the interception technology, on live customer data. There was no customer permission obtained for these. I have asked the official BT spokesman to confirm or deny this. Answers - i hope - on the beta.bt groups. A CTO, by definition, would have knowledge of such trials, and I would expect that he would be the senior authoriser. In the event of a criminal prosecution for last year's events, it looks possible that he would be personally held to account. (Thanks to The Register for pointing out the CTO connection) John L John |
|
#2
|
|||
|
|||
|
John wrote:
> It is starting to look as if BT's involvement in this was greater that > it has been reported so far, possibly to the extent of being the primary > development partner. > > Statis Scleparis, who was BT Retail's Chief Technology Officer between > 2006 and 2007 is now the CTO in Phorm. Evidence here - > http://www.phorm.com/about/exec_scleparis.php > > It appears that Scleparis may have been the Retail CTO during last > year's covert trials of the interception technology, on live customer > data. There was no customer permission obtained for these. > > I have asked the official BT spokesman to confirm or deny this. > Answers - i hope - on the beta.bt groups. > > A CTO, by definition, would have knowledge of such trials, and I would > expect that he would be the senior authoriser. In the event of a > criminal prosecution for last year's events, it looks possible that he > would be personally held to account. > > (Thanks to The Register for pointing out the CTO connection) > > John L I was browsing Phorm's site the other day and noticed myself that Scleparis had come over from BT. It's developing into a very juicy story indeed - it's been quite entertaining watching BT's PR guys running backwards at high speed. -- Rev Robert M Jones, Wimborne Baptist Church, UK Hub forums here: http://hubbub.labs.bt.com/ Hub info & forum here: http://www.frequencycast.co.uk/homehub.html http://www.wimborne-baptist.org.uk Free trial of Mailwasher Pro - effective email spam filter - (commission goes to our partners in Bulgaria) http://fta.firetrust.com/index.cgi?id=420 |
|
#3
|
|||
|
|||
|
John wrote:
> It is starting to look as if BT's involvement in this was greater that > it has been reported so far, possibly to the extent of being the primary > development partner. > > Statis Scleparis, who was BT Retail's Chief Technology Officer between > 2006 and 2007 is now the CTO in Phorm. Evidence here - > http://www.phorm.com/about/exec_scleparis.php > > It appears that Scleparis may have been the Retail CTO during last > year's covert trials of the interception technology, on live customer > data. There was no customer permission obtained for these. > > I have asked the official BT spokesman to confirm or deny this. > Answers - i hope - on the beta.bt groups. > > A CTO, by definition, would have knowledge of such trials, and I would > expect that he would be the senior authoriser. In the event of a > criminal prosecution for last year's events, it looks possible that he > would be personally held to account. > > (Thanks to The Register for pointing out the CTO connection) > > John L 80/20 report on Phorm is now published http://blogs.guardian.co.uk/technolo...20final%20.pdf Phorm have been quoting this as supportive, although two of the members of the group have come out elsewhere with concerns about illegality. The report itself seems to raise a lot of questions both about the system itself, AND the way ISP's are expected to be communicating with their customers. -- Rev Robert M Jones, Wimborne Baptist Church, UK Hub forums here: http://hubbub.labs.bt.com/ Hub info & forum here: http://www.frequencycast.co.uk/homehub.html http://www.wimborne-baptist.org.uk Free trial of Mailwasher Pro - effective email spam filter - (commission goes to our partners in Bulgaria) http://fta.firetrust.com/index.cgi?id=420 |
|
#4
|
|||
|
|||
|
Robert M Jones wrote:
> > 80/20 report on Phorm is now published > http://blogs.guardian.co.uk/technolo...20final%20.pdf > > Phorm have been quoting this as supportive, although two of the members > of the group have come out elsewhere with concerns about illegality. > > The report itself seems to raise a lot of questions both about the > system itself, AND the way ISP's are expected to be communicating with > their customers. > Here's some of the 80/20 report on Phorm published in the last day or so. There is some interesting stuff here that ISP's in particular need to take note of, in relation to keeping customers fully informed, implementing "opt-IN" policies, and also having regular reminders and regular repeated consent about opt-in/opt-out because of the number of computers/users in households. Questions asked about monitoring of less common, less secure email sites (not using https.) Some disagreement between Home Office and 80/20 about basic position under RIPA! A reminder - the 14 page report is available in full here: http://blogs.guardian.co.uk/technolo...20final%20.pdf Some extracts:- despite our positive findings regarding Phorm’s approach to privacy protection we are disappointed that the company has not benefited from an earlier implementation of a PIA. While we are encouraged that Ernst & Young were engaged to perform a privacy examination, the full scope and influence of an “early intervention” PIA has not been possible. At this late stage of product development it will not be possible to fully exploit the value of a PIA. We broadly agree with the positive findings of the 2007 Ernst & Young privacy examination, but remain concerned that the scope of that report was based almost exclusively on conditions applying to the US privacy environment. Public sensitivities, regulatory conditions and other factors vary substantially according to geographical location. We believe it will be crucial to devise a system based on both transparency and embedded technological safeguards to provide assurance that Phorm Technology does not fall victim to the level of function creep evident in other technologies. In our view, Phorm should ensure that ISPs clearly communicate with their users about the issues involved in Phorm Technology surveillance, and actively and regularly pursue users' consent. We believe this approach may be crucial to mitigating potential concerns about surveillance. Communications surveillance laws at the very least require consent to be re-affirmed at regular intervals, particularly as multiple users may make use of a single Internet connection and machine. Phorm's privacy policy responsibly notes that Phorm may disclose information to third parties under 'legal requirements'. Considering how legal protections vary by country, far more information is required for users to ensure their confidence in the data processing. Although the PIA process takes the Data Protection Act and other relevant laws into account, it does not focus exclusively on them. A complementary audit process is needed to ensure that the project is legally compliant. That process can begin early, but cannot be finalised until late in the project lifecycle, when the design is complete. Phorm liaised with the Home Office to assess whether its system could infringe the UK law that regulates communications surveillance. The Home Office concluded that Phorm's system is consistent with the Regulation of Investigatory Powers Act and does not intercept communications. While this conclusion is a fair interpretation of Phorm and the system's capabilities, communications monitoring still takes place. Even if the Home Office's conclusions were appropriate and relevant, it would mean that if an ISP or any government wished to conduct similar monitoring of communications for segmentation purposes, albeit with consent of the user, then they may indeed do so and yet still be compliant with UK law. This could indeed give rise to a worrying situation. In its assessment, the Home Office compares targeted online advertising with email/spam filtering. This was a similar line of argument pursued by Google in its Gmail advertising service: the content of messages are already being processed by ISPs to assess whether they are spam, therefore analysing content for advertising purposes is no different. The key difference, as argued by many privacy experts, is that processing communications to remove inconveniences (e.g. spam) is not invasive because it is intentionally not passing judgment on the user. Processing communications to categorise individuals, or to pass judgment on the consumer, is a privacy interference. Phorm must ensure that ISPs clearly communicate with their users about the issues involved in this 'surveillance', and actively and regularly pursue users' consent. This is the only way to mitigate concerns about surveillance. Ideally some form of black-list of sites should be included, or a white-list with clear exclusion processing. For instance, even though Phorm's system excludes forms, and therefore would exclude content from sites where an individual is drafting an email, and also excludes https traffic which therefore excludes many webmail service providers, users would need strong assurance that the process through which they read emails (on less-secure platforms) is not also being monitored. Can user-sensitive URLs be excluded? While Phorm is careful to note that HTTPs pages are processed this is perhaps more a matter of an inability to gain access to the content of these pages because they are encrypted. Are https-requests not logged at all? That is, 1080-requests tend to be from servers where users have an existing relationship, e.g. their banks, travel agents, mail providers, and places where the user shops. If this information was to be logged by an ISP this would make users feel spied upon because their ISP would know which services he or she makes use of. Phorm must ensure that it is not using information about these sites in any way, e.g. URL data. We are aware that only widely-viewed pages will be used, possibly to limit profiling to highly specific user data. This is certainly a positive development. Phorm must communicate this fact to end-users. Similarly, users need to be informed explicitly about the constitution of channel information. If not carefully explained, users may worry that channel information, depending on the level of data granularity, is in itself personal or sensitive information. For instance, if a channel is able to discern that a user banks online, uses a non-online insurance company, this could be seen as personal information particularly where the user's bank and insurance company could be known to the profiler. Therefore clearer information is required about how the profile is developed and how this information is combined with the channels. Consent and Participation To adhere to the highest principles of data protection, any system that processes personal information must require consent on an opt-in basis. As Phorm's system involves a form of communications surveillance then optimal protections would involve opting-in. The market default for cookie-based consent systems is opt-out however. Phorm's chosen implementation matches market practices. Phorm goes some way to mitigate this concern by creating a website for opting-out and encourages partners to remind users about opt-out rights. We would like to hear more about this form of 'encouragement' to clarify the role of Partners in ensuring privacy practices are pushed to the highest level possible. Communications surveillance laws at the very least require consent to be re-affirmed at regular intervals particularly as multiple users may make use of a single Internet connection and machine. Further challenges exist and clarifications are required. - If a user blocks all cookies (or manages cookies on an opt-in basis), these users will have to be informed about how their traffic is managed by the Phorm system. That is, if there is no cookie present does the traffic still get processed? It is important to be clear to users that if they choose not to participate in the system at all then their traffic is not being processed. - If a user regularly deletes cookies then this would result in that user being monitored again. Ideally a user would be able to notify his or her ISP that he or she is uninterested in participating in the advertising scheme altogether and this would result in a permanent non-processing of Internet traffic. Is such an implementation possible? - With limited information about the channels and profiles, a user may be concerned about seeing which 'channel' they have been linked to and the means through which this decision was made. Phorm must develop educational materials for users to understand this process. Similarly, Phorm must explain how many possible channels there are in case users are worried about being segmented in great detail. dentity, Traceability, and Security Phorm is very careful in the design of its system and in its public information avoid processing personally identifiable information. Phorm's system itself does not process IP addresses and promises that it does not link back to ISP's subscriber databases. Concerns remain, however: - Can cookies lead back to users in any way? Of course it is merely a unique identifier but a unique identifier can still be linked to individuals. Can an external attacker gain access to the required information to re-link the individual and the UID? Even if this was possible, what potential gain could there be for an attacker? - Phorm's privacy policy responsibly notes that Phorm may disclose information to third parties under 'legal requirements'. Considering how legal protections vary by country, far more information is required for users to ensure their confidence in the data processing. We would be interested to know what kind of information Phorm and its system actually holds that may be of interest to third parties. This of course refers back to the linkability issue: if the profile nor the advertising information not linkable to the individual then of what use would such data serve to third parties such as law enforcement authorities? - Linked to the above two point, if there was a malicious insider, with complete access to all the traffic and transactions, could re-identification take place? Or could any level of traffic analysis generate persona data about the user, the types of advertisements served, and the user's IP address? Although the security statement in the privacy policy is a responsible statement, Phorm's security policy and security processes should be audited regularly. -- Rev Robert M Jones, Wimborne Baptist Church, UK Hub forums here: http://hubbub.labs.bt.com/ Hub info & forum here: http://www.frequencycast.co.uk/homehub.html http://www.wimborne-baptist.org.uk Free trial of Mailwasher Pro - effective email spam filter - (commission goes to our partners in Bulgaria) http://fta.firetrust.com/index.cgi?id=420 |
|
#5
|
|||
|
|||
|
John wrote:
> It is starting to look as if BT's involvement in this was greater that > it has been reported so far, possibly to the extent of being the primary > development partner. > > Statis Scleparis, who was BT Retail's Chief Technology Officer between > 2006 and 2007 is now the CTO in Phorm. Evidence here - > http://www.phorm.com/about/exec_scleparis.php > > It appears that Scleparis may have been the Retail CTO during last > year's covert trials of the interception technology, on live customer > data. There was no customer permission obtained for these. > > I have asked the official BT spokesman to confirm or deny this. > Answers - i hope - on the beta.bt groups. > > A CTO, by definition, would have knowledge of such trials, and I would > expect that he would be the senior authoriser. In the event of a > criminal prosecution for last year's events, it looks possible that he > would be personally held to account. > > (Thanks to The Register for pointing out the CTO connection) > > John L More bad news for Phorm The Guardian, with the largest newspaper online site in the UK, has just pulled the plug on their relationship with Phorm. http://www.theregister.co.uk/2008/03...n_phorm_uturn/ Quote from Register article: <<...advertising manager Simon Kilby revealed the retreat: It is true that we have had conversations with them [Phorm] regarding their services but we have concluded at this time that we do not want to be part of the network. Our decision was in no small part down to the conversations we had internally about how this product sits with the values of our company. I hope you appreciate that the quality of the Guardian's editorial is funded by our advertising sales operation and it is our duty to keep abreast of all developments in this sector. In this instance, however, I agree with you that this is not something that we should be partnering.>> -- Rev Robert M Jones, Wimborne Baptist Church, UK Hub forums here: http://hubbub.labs.bt.com/ Hub info & forum here: http://www.frequencycast.co.uk/homehub.html http://www.wimborne-baptist.org.uk Free trial of Mailwasher Pro - effective email spam filter - (commission goes to our partners in Bulgaria) http://fta.firetrust.com/index.cgi?id=420 |
![]() |
| Tags |
| covert, links, phorm, trials, year |
| Thread Tools | |
| Display Modes | |
|
|