Is Microsoft “snooping” on Skype text conversations, or merely protecting users from malware URLs?
German publisher Heise Online has given that question prominence with the accusation that Redmond is snooping, as the result of receiving return visits from Microsoft IP addresses if they send HTTPS URLs through Skype text chats.
In essence, untangling the syntax that Google Translate applies to German – I think it passes through a couple of other languages on the way to English – Heise reports that if two Skype users send a URL through text chat, they get an “unannounced visit from Redmond”.
The issue was raised by a reader, the report says, whose network protection took the tap from a Microsoft IP address as a replay attack.
To test the report, the researchers sent URLs – one of them including HTTPS credentials – via Skype, and then watched the logs of the servers identified in the chat messages. The servers then received visits from the Redmond machine (IP address 22.214.171.124) using the same URLs (including the credentials, Heise claims).
Redmond's response to Heise's inquiry was that it is scanning URLs to make sure that users aren't (accidentally or stupidly) passing around malware or phishing links. The publisher retorts that instead of using an HTTP GET, Microsoft is calling the URLs using HTTP REQUESTS HEAD. As a result, it says, Microsoft isn't actually reviewing the pages it's requesting.
The Register isn't so sure that the HEAD request is useless: it could offer a quick and automated way to see if a page's header matched a signature for known malicious pages. However, repeating requests that include sign-on credentials would appear, at first blush, to be a risky behaviour. ®