Three questions that could put out Amazon's Fire
US politician probes Bezos' Silk Road
A US congressman is pushing Amazon for details of its cloud-based browsing, Silk, specifically asking what data the company is gathering and how it intends to make use of it.
In an open letter (2-page PDF/263KB, short and to the point) Congressman Edward Markey asks Amazon's CEO Jeff Bezos specifically what information is being collected by Silk, how Amazon intends to make use of it, and how the company will go about ensuring users have given explicit permission to have their behaviour monitored in this way.
That last point intimates that such permission should be explicitly requested, while Amazon was probably hoping that the usual user assent to unread Terms & Conditions would suffice.
Amazon's new Kindle Fire is an Android tablet that relies on Amazon's cloud to pre-process web content, a facility Amazon has titled "Silk". Users can opt out of Silk, but by default every click will be routed through Amazon's servers.
This is nothing new: mobile browsers such as Opera, Skyfire and Bolt all do the same thing – grabbing web content and pre-processing to optimise and compress the content prior to delivery – and all three have access to the same kind of information that Amazon Fire users will be sharing.
But none of the above have the reach, or ambition, of Amazon, which has made it clear that user profiling is very much part of its business plan. Our own Andrew Orlowski recently compared Amazon Silk to Phorm, the intercept-and-track service trialled in the UK by BT and still being deployed elsewhere, pointing out how both have the potential to invade users' privacy pretty equally.
When Phorm started collecting data, there was uproar. When Amazon announced the same thing, it seemed as if no one cared.
But at least one US congressman does, it seems, and he is expecting Jeff Bezos to explain himself by 4 November. We'll keep an eye on the matter, and look forward to sharing what comes back. ®
Sponsored: Benefits from the lessons learned in HPC