Even seemingly safe web addresses are rife with attack code aiming at vulnerable clients, according to a new study from the Honeynet Project. The study also found that methods such as blacklists can be surprisingly successful in stopping client-side attacks.

Attackers are increasingly turning to end-user systems as a way around the antivirus and firewall systems that are increasingly blocking access to traditional attack routes, according to the researchers, who hail from the US, Germany and New Zealand.

"The 'black hats' are turning to easier, unprotected attack paths to place their malware onto the end-user’s machine," they said in the study, called "Know Your Enemy: Malicious Web Servers."

The researchers, using a "high-interaction" client honeypot called Capture-HPC developed by the Victoria University of Wellington, analysed more than 300,000 addresses from around 150,000 hosts.

The study looked at various site categories, including adult, music, news, "warez," defaced, spam and addresses designed to grab traffic from users who mistype common web addresses. While some categories were more likely to contain malicious addresses than others, all contained malicious addresses, the report said.

"As in real life, some 'neighborhoods' are more risky than others, but even users that stay clear of these areas can be victimised," the report said. "Any user accessing the web is at risk."

Users can be led to malicious sites via links, typing in an address manually, mistyping an address or following search-engine results, the study said.

Safeguards

These results only confirm what security researchers have been saying for some time now. But the study also analysed the effectiveness of safeguards against such infections in some detail.

The research showed that blacklists, if regularly updated, can be a surprisingly effective way of blocking malicious addresses.

The researchers also recommended regular patching, but this may not always be straightforward, since the study found a prevalence of attacks against plug-ins and non-browser applications. "Attacks also target applications that one might have not think about patching, such as Winzip," the study said.

Another technique that can block attacks would be to use a less popular browser, such as Opera, the study found. "Despite the existence of vulnerabilities, this browser didn’t seem to be a target," the study said.

The data used as the basis for the study has been made available on the Honeynet Project's website.