I Have a Security
Dream"> It was year 2000 when I wrote, way, way too early, that ISPs were the future for security for consumers. But I had a different idea of how ISPs would provide security: I saw it entirely as network-based feature. Almost all malware these days comes through the Internet to the user, so it seemed to me that the ISP should be scanning all channels of communication (in effect, all TCP ports) going to or coming from the user.Its ambitious in terms of the computing power it would require, but its easy to assert that such computing power will be readily available in the future. Here we are in 2005, and my prediction is still Buck Rogers stuff. Oh, there are some ISPs who do host-based scanning of mail for malware and spam, but thats less than half the problem. If ISPs had a network firewall of sorts for their users, they could stop the next Sasser or Blaster from affecting their networks. Click here to read more from columnist Jim Rapoza about ISP security. Isnt that an appealing idea? Its going to happen in the future. I hope I live long enough to see it. In the meantime, these new solutions are based on the distribution of traditional client-based security software, and thats good enough. Well, basically it is. There are interim measures that can improve the situation, and these you might see developing in the next few years. CA allows the ISP to customize the software both aesthetically and with policies; so for instance, the software could be set to block, by default, all outbound connections on port 25 (SMTP)other than those to the ISPs mail server. If this is a problem for some users, they can change the policy. The ISP cant go so far as to manage the client systems software by changing policies remotely, except perhaps by pushing updates to the user. This is the sort of capability one finds in corporate versions of client-based security software, and is removed from consumer versions partly because there is nobody managing the system and partly to protect the higher price of the corporate edition. It might also be useful for the ISP to check whether the client systems software is running and current. After all, new threats often try to disable security software. One way ISPs could do this is by implementing a dream I have: network access control for ISP networks. (Yes, I know, I have strange dreams.) This is a system implemented by a variety of companies (Sygate, for example) for setting configuration requirements for a system before it is allowed on the network. These can include security requirements such as running anti-virus with current definitions. Network access control is great stuff, but its only a corporate feature now. I think that eventually one could imagine ISPs offering a special "locked-down" area with access control and other network controls and sell it as a "super secure" ISP area. It would probably have to be more expensive because it would need more hand-holding, but some users would see value in it. Its good to see ISPs trying to be part of the solution to our security problems. There are still others, unmentioned in this article, who are most definitely part of the problem. They still dont care what happens on their users computers and the effect it has on the network and Internet beyond. I can see them getting displaced in the market by ISPs that do care. At least I can dream about it. Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983. He can be reached at firstname.lastname@example.org. Check out eWEEK.coms for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzers Weblog.
All users would be protected, not just those who have the security software running on their computers.