Internet 'phishing' scams getting more devious

WASHINGTON - internet "phishing" scams are becoming more difficult to detect as criminals develop new ways to trick consumers into revealing passwords, bank account numbers and other sensitive information, security experts say.

Scam artists posed as banks and other legitimate businesses in thousands of phishing attacks last year, sending out millions of "spam" emails with subject lines like "account update needed" that pointed to fraudulent websites.

These attacks now increasingly use worms and spyware to divert consumers to fraudulent sites without their knowledge, experts say.

"If you think of phishers initially as petty thieves, now they're more like an organised crime unit," said Paris Trudeau, senior product manager for internet-security firm SurfControl.

Phishing attacks have reached 57 million US adults and compromised at least 122 well-known brands so far, according to several estimates.

At the end of 2004 nearly half of these attacks contained some sort of spyware or other malicious code, Trudeau said.

One attack, first documented last month by the Danish security firm Secunia, misdirects Web surfers by modifying a little-known directory in Microsoft Windows machines called a host file. When an internet user types a Web address into a browser, he is directed instead to a fraudulent site.

This technique has shown up in attacks spoofing several South American banks, said Scott Chasin, chief technical officer of the security firm MX Logic.

The convergence of all of these threats means "we can expect to see some large attacks in the near term," he said.

Another more ambitious attack targets the domain-name servers that serve as virtual telephone books, matching domain names with numerical addresses given to each computer on the internet.

If one of those computers is compromised, internet users who type in "www.bankofamerica.com" could be directed to a look-alike site run by identity thieves.

Domain-name servers are tougher to crack, as they are typically run by businesses rather than home users, but hackers can find a way in by posing as a company's tech-support department and asking new employees for their passwords, Trudeau said.

Domain-name hijacking is suspected in incidents involving Google.com, Amazon.com, eBay Germany and HSBC Bank of Brazil, Chasin said.

Even straightforward phishing attacks are getting more sophisticated. Spelling errors and mangled Web addresses made early scams easy to spot, but scam artists now commonly include legitimate-looking links within their Web addresses, said Kate Trower, associate product manager of protection software for EarthLink Inc.

Consumers who click on links like www.citibank.com in these messages are directed to a fraudulent Web address buried in the message's technical code, she said.

MasterCard International has caught at least 10 phishing scams involving www.mastercard.com over the past two months, said Sergio Pinon, senior vice president of security and risk services.

Consumers can protect themselves with software that screens out viruses, spyware and spam. But online businesses will have to take steps as well, perhaps by issuing customers a physical token containing a changing password, Chasin said.

Internet engineers should also figure out a way to authenticate Web addresses, much as they are currently figuring out how to make sure email addresses are legitimate, he said.

- REUTERS