Spam Filters Gone Wild

Spam Filters Gone Wild; Spate of Incidents at Verizon, AOL Point to Growing Problem Of Blocking Legitimate Email

By JESSICA E. VASCELLARO May 3, 2006

Internet companies are taking more aggressive steps to stop the flow of unwanted email. In a significant number of cases, though, consumers complain that the efforts increasingly are blocking the good along with the bad.

Possibly millions of AOL members were temporarily unable to receive some mail from Google Inc.'s Gmail users last week after AOL held up messages from some new Gmail servers over concerns it might be spam. An AOL software update recently resulted in a stoppage of mail that mentioned at least 60 Internet addresses. An update of Verizon Communication Inc.'s spam filters recently sparked widespread complaints from consumers who were unable to receive and send messages.

The companies blamed the problems on software glitches or communication failures and often fixed them within hours. Tight precautions are necessary, the companies say, since spam can threaten online security and safety -- a more serious problem than the nuisance of a few missed messages. But others say the incidents are a troubling sign that new antispam measures may be going to far, contributing to everything from lost real-estate deals and blocked banking transactions to bruised relationships caused by unreturned emails that never got through to friends in the first place.

[TELECOM Digest Editor's Note: As I discuseed in this column yesterday, the _real villains_ -- the spammer-scammers -- do not care either way, and the people who lean toward caring are busy squabbling with each other over which methods to use. If one has to choose between 'online safety and security' (the viewpoint of many companies) and 'a few missed messages' (which netizens seem to care about) I would choose to try and insure that no _legitimate_ messages were lost. But there is a third choice, which some netizens -- I call them 'enablers' refuse to consider: severe punishment of spammers; they opt instead to continually try to refine their 'spam filters', not worrying all that much about 'a few messages'. The enablers have one excuse after another why nothing will work except for their filtering schemes, and as we see in this message and others, even that does not work all that well. Maybe they can learn by extremes: I suggested yesterday that if lots of mail servers were constantly jammed with trash domain names (for our pretection; valid email addresses in the text area of email) and a notation that spammer-scammers 1, 2, 3, 4 had made these efforts necessary (let's all give a warm greeting to them!) they might possibly begin to take the hint; then again, sadly, maybe not. PAT]