Experts: Google AdWords needs policing

Google could avoid future malware attacks carried out using advertisements posted on its Web sites if the company more thoroughly investigated customers of its AdWords system, according to security and legal experts.

On April 25, researchers with security software maker Exploit Prevention Labs announced that they had uncovered hard evidence that malware distributors were using advertisements placed via Google’s automated AdWords system to infect unsuspecting end-users with virus code.

According to Roger Thompson, chief technology officer at Exploit, based in New Kingstown, Pa., the malware brokers used fraudulent advertisements for legitimate organizations such as the Better Business Bureau to trick users into clicking on the links.

When someone clicked such a link, the ad would redirect their browser through URLs that attempted to automatically download virus programs onto their computers before passing them along to the actual sites that were advertised. The system works so quickly that end-users do not notice that their browsers have traveled through the intermediary sites, making it almost impossible to detect the attacks, Thompson said.

Using AdWords, any company or individual can bid on specific terms to have their ads show up next to Web search results for the specific words they have purchased. People who bid the most money for any particular term or set of words receive the highest ad placement, or so-called sponsored links, next to Google’s Web search results.

Thompson said that his team discovered a number of the ad-borne threats built to show up alongside results for a range of search terms, including queries for information on business conferences and auto shows. Many others involved terms related to business organizations.

The company said it only found the threats after one of its customers noticed the suspicious activity using the software maker’s secure Web surfing application.

When the individual entered the phrase “how to start a business” into Google, the top-ranked sponsored search listing advertised AllBusiness.com, a legitimate organization, but the hyperlink it offered led to a site that attempted to install a password-stealing keystroke logging attack on the user’s PC, according to Exploit.

The security company traced the malware distribution scheme back to Smarttracker.org, a domain name registered in early April that began advertising on Google soon thereafter. The fraudulent advertisements all funneled users through infected Smarttracker URLs, Thompson said.

Google officials said that the search company canceled the affected ads after it was informed of the situation. Media representatives with Google said the firm is aggressively working to eliminate such problems.

“We actively work to detect and remove sites that serve malware to our users both in our ad network and in our search results,” the company said in a statement. “We have manual and automatic processes in place to detect and enforce these policies; we also encourage our advertisers to contact Google directly if they have concerns or detect suspicious malware.”

Thompson and other security experts claim, however, that the problem exists because Google does not sufficiently police its advertisers.

“Google says they are doing the best that they can, but their business model is to take as much money as they can for advertisements. No matter how much due diligence they do, it’s a difficult position to be in, but clearly they are not doing enough,” Thompson said. “If they don’t do a better job of vetting their customers, we will see this sort of thing happening again and again.”

Thompson pointed out that his company only caught on to the problem because the attackers were using malware exploits that were discovered and fed into its LinkScanner software. If the malware brokers had used new code and varied their attacks, it is likely that the hacks would have carried on unnoticed for a longer period of time, he said.

The hacks involved were of danger to users of Microsoft Windows software who had not updated their machines with current security patches.

“We started seeing links back to Smarttracker but didn’t figure out how they were hooking people until someone happened to stumble upon the ads in their normal Web surfing. Nobody knows where to look for this stuff without that kind of luck,” Thompson said. “The domains these people can use are infinite, but the exploits are recognizable, so that’s why we caught them at all.”

One person unsurprised by the unfolding infected Google ad scenario is Ben Edelman, a well-known expert on legal issues affecting the Internet and online advertising, who is employed as an assistant professor at Harvard Business School.

Edelman has been following unsavory tactics carried out by malware and adware distributors for years, and said that he had observed similar activity on Google as far back as one year ago.

In addition to allowing hackers to pervert AdWords by using the system to trick end-users into consuming malware, the Harvard-trained lawyer pointed out that Google also allows people to post dishonest advertisements that offer paid downloads of free software applications, such as the Skype Internet-calling tool.

“The big problem is that Google sells ads to anyone, without completing any due diligence to determine who they are or that the content they are advertising is legitimate,” Edelman said. “We’ve been writing about this problem for years and nothing has been done about it. Apparently anyone who pays the bills is good enough for Google.”

The expert noted that people trust Google because it is a large, successful online company. The infected ads may even show up on the company’s Gmail Web mail client, Edelman said, which requires users to register with the site and use passwords to access their accounts, creating another false level of protection from threats.

“People treat sponsored results as safe because they believe that Google has filtered out the bad stuff, which they should, because legitimate publishers such as print or broadcast companies do that,” Edelman said. “It’s highly likely that these malicious ads appear throughout the Google network, including in Gmail, and they may also show up on sites like AOL and Ask.com that are advertising syndicates.”

Edelman said that the only answer to the problem, which he said affects most search engines, not just Google, is for the involved companies to better police their advertisers, or for lawmakers to force them to do so.

“I think a lot of people might favor legislation to make it clear that search engines have the same responsibility to verify ads that print publications have today,” Edelman said. “It’s the search engine companies’ responsibility to do a better job of protecting their users.”