A plethora of antispam products exist; some integrate with the client, some offer hosted services, and some run on a server. All the products have one thing in common, though: They're largely reactive. A year ago, who could have predicted the extent to which spammers would falsify subject lines and use messages made up entirely of images? And who knows what the spammers will be up to next year? As spammers change their tactics, antispam tools must add new filters, and you must take the time to update your protection if you want to continue to block spam messages. Some promising new developments in the war on spam are in the works, though, and they raise the possibility that more proactive approaches to spam fighting might soon be available.
First is the announcement that AOL, MSN, and Yahoo! have joined forces to fight spam by beefing up enforcement of consumer protection laws, hardening their services to make them more difficult for spammers to use, and providing better filters. These three companies have historically been unable to agree on anything, but perhaps fighting spam is a goal important enough to cause them to form an alliance that will bear fruit.
Second is the rise of some new methods for detecting spam. Bayesian analysis is a fascinating branch of statistics that can help a filter "learn" how to tell which messages are spam and which aren't. Products that implement Bayesian filtering require training (i.e., you must initially input a certain amount of data) but tend to be highly accurate once fed a large enough corpus of data, and extending their training to recognize new classes of spam is as simple as feeding messages of those types into the filter. In a similar vein, services that let one client report a message as spam so that other clients can block the same (or similar) messages are gaining in popularity. The efficacy of these services depends on the number of mailboxes they filter, so such services will continue to grow in effectiveness as they catch on with users.
Third is the rise of new legislation (although generally I'm not a big fan of government-mandated solutions, especially when the problem relates to the global Internet, I make an exception in the case of spam). A proposed law, for example, would require spammers to include a subject-line tag on their messages and would offer a bounty to people who successfully track down spammers who fail to comply. In this scheme, the US Federal Trade Commission (FTC) would be able to sue spammers who break the law, and whoever does the legwork necessary to track down such criminals would get 20 percent of the fine. Not a bad deal, although the law still has some rough edges.
Do these advances mean we can relax? Unfortunately, not yet. Even with such measures in effect, you still need to be vigilant about filtering spam on your servers, both to protect your storage and messaging resources and to spare your users unnecessary hassle. However, I hope that at least one of these new developments will make life easier for us in the long run. In the meantime, keep updating those filters!