Software giant Microsoft has u-turned on last year's controversial proposal to establish a new Global Collective Defence (here), which among other things also called upon broadband ISPs to restrict internet access for customers with computers that had become infected by Malware (Viruses, Trojans etc.).

Scott Charney, MS's Corporate Vice President of Trustworthy Computing, told this year's RSA Security Conference that his "thinking [had] evolved" since then, before explaining how he now recognised the "many flaws with that model".


At the time it didn't take a genius to work out that the affected computers would most likely be those with Microsoft Windows installed; somewhat of an own goal for the software giant. Likewise most infections do not completely destroy a computer and the affected individuals might still need their internet access to find a solution.

Instead of that Microsoft has now proposed an Internet Health Certificate for computer systems. The idea is that online banks, websites or other services could check the user's computer for this certificate and decide whether or not to give them access. For example, a user might be stopped completely or only given partial access to their bank if the anti-virus software was not up-to-date.


Naturally there's a fatal flaw. One of the first things virus authors will do is to write code that makes the computer appear secure :glee: and as a result the certificate would instantly become irrelevant.

Meanwhile Kaspersky Lab's, a popular Anti-Virus firm, yesterday called on worldwide governments and broadband ISPs to consider tough new measures that would allow ISPs, anti-virus firms or the police to "forcibly remove malicious code" from infected computers (here).

As both we and Charney have recently pointed out, some ISPs do this already but it's extremely difficult and costly to accurately identify infected computers. Likewise, even when you do, only a tiny fraction ever appears to be caught.