Undetectable ‘son of cookie’ system wins grant
It knows where you're surfing, right now
The developers of a 'son of cookie' web monitoring system have received a Proof of Concept grant from Scottish Enterprise to commercialise the system. Their non-cookie based web monitoring software does not (as indeed the name suggests) rely on cookies, but instead is intended to replace them with something far more powerful.
It has, as the features list makes clear, great privacy-invading potential. The "sensors" it uses:
- can be individually customised for any web visitor;
- can collect information rather than return pre-downloaded data.
- can be reconfigured remotely;
- are difficult to detect and delete;
- can be used to block access to sites, documents, data, emails, etc., based on content,
- can be preferentially customised for each user.
So the system can provide highly detailed tracking information for market research purposes, and "is also suitable for Internet and general computer surveillance on behalf of commercial organisations, governmental bodies and educational establishments alike. It can enable tracking of visitors to any website worldwide, and help to address Internet crime."
But dont worry, because: "The development of appropriate safeguards to prevent misuse of the technology in these contexts will be developed in parallel to the technology itself. These are a critical part of this project, and will include a modular approach to allow exclusion of technical capability and prevention of sensor re-configuration."
How, though, does it work? If it works, that is. This isn't entirely clear from the Scottish Enterprise announcement or the Strathclyde University press release announcing the award of the grant, but a Strathclyde's Department of Electronic and Electrical Engineering has a clipping from Business A.M. that gives more details.
The "sensor" program can monitor keystrokes and "the whole range of a user's internet usage," and it "can also be altered remotely, allowing it to be fine-tuned to the owner's or user's particular requirements." We trust you're as interested in that differentiation between 'owner' and 'user' as we are. The piece also suggest that companies gathering data could get their customers to use the software in exchange for a payment or discount on purchases, so clearly there's a component of the software that has to be run locally.
So you can just say no, right? Up to a point, we'd hazard. If it were absolutely clear that users would be warned and given the ability to refuse it, then there would really be no need for it to be pointed out that it is "difficult to detect and delete." So you could speculate about it coming hardwired and unannounced in, say, a bank's client software, being rolled out for security and monitoring reasons to all of the clients in a company network, or being sent to Microsoft Outlook users "in order to have your advice."
We hope the "appropriate safeguards" will be sufficient to take care of that last one, but if it's as powerful and unobtrusive as they suggest, it's difficult to see how abuses can be blocked. ®
Sponsored: DevOps and continuous delivery