'Nightclub bouncer' phishing scam found by researchers
Cyber-criminals have invented a cunning new method of targeting victims by developing a system that behaves like a bouncer at an exclusive nightclub.
Security firm RSA revealed how attackers assigned targets with a unique ID, meaning the scam could be aimed at specific people.
If a person's ID was not on the list, their computer would not be affected.
RSA said the advanced threat posed a new "detection challenge" to the security industry.
In a blog post, RSA cybercrime specialist Limor Kessem wrote: "As we adapt and improve our detection systems, we are reminded that in the never-ending cat-and-mouse game, only the nimble will survive."
The so-called Bouncer Phishing Kit targets preset lists of email addresses. For each target, a unique ID is automatically generated, creating a unique web address for the user to click on.
If someone has an ID that does not match the list of intended targets, they will simply be presented with a 404 Error page, and will be unharmed.
If, however, a person is one of the unfortunate ones, the same page will instead spring into life as an "attack page" ready to steal user credentials.
Using this method means attackers can harvest data from certain groups of users, rather than having to sift through large amounts of data.
For example, the Bouncer Phishing Kit could be used to gather personal details on people in one particular country.
"It holds this [bouncer] moniker because much like many high-profile night-time hotspots - if your name is not on the list, you're staying out," Mr Kessem said.
"Traditional phishers like to cast as wide of a net as possible.
"But with this tactic the phisher is laser-focusing the campaign in an effort to collect only the most pertinent credentials for his purposes.
"Keeping out uninvited guests also means avoiding security companies and prompt take-downs of such attacks."
Phishing is a growing problem for internet users. According to RSA's data, attacks of this type were up 59% in 2012 compared with the previous year, and cost the global economy $1.5bn (£940m).