Something fishy about Google Chrome’s Safe Browsing API, lab says

From the start, Google’s Safe Browsing API was designed to spot malicious web pages so users wouldn’t get trapped in them. Google identifies these sites through its own algorithms and user notification.

Google Chrome isn’t the only browser to do this. FireFox and Safari rely on the lists made available in the Safe Browsing API, and Microsoft has its Application Reputation with Internet Explorer, which essentially does the same thing.

This week, NSS Labs, a firm that specializes in the testing of security systems, found something in its monitoring that just didn’t feel right.

According to NSS Labs, during the most recent period of testing, Nov. 21, 2011 through Jan. 5, 2011, they observed what appears to be a significant change in malicious website protection when contrasted with historical data. According to their report, “Did Google Pull a Fast One on Firefox and Safari Users?“, Chrome’s protection rate rose to more than 50 percent before falling back down to 20 percent, while at the same time the Firefox and Safari block rate remained stuck at 2 percent and then suddenly jumped to 7 percent on the same day Chrome’s protection precipitously dropped.

The types of attacks NSS Labs evaluated during this period are what it calls “socially engineered malware,” or malware that is downloaded by the user from the web. The lab will be testing so-called drive-by download attacks in a later report.

“Google has made very public statements that they don’t withhold any data from their Safe Browsing API, so what could explain the results?” asks Vikram Phatak, chief technology officer at NSS Labs.

Perhaps it’s the undocumented functionality NSS Labs believes Google has integrated into Chrome, but not Firefox or Safari.

Google strongly denies it’s holding back anything from the API. In his blog, New SafeBrowsing Backend, Mozilla and Mobile Firefox developer Gian-Carlo Pascutto at first wrote that Firefox does not have permission to use the download protection list in the Safe Browsing API.

That statement has since been redacted following a response from Google, a response that highlights perhaps a deeper concern: privacy.

“We have offered the new Safe Browsing features to Mozilla in the past, so to say that we are holding back this functionality is inaccurate. From our conversations, our understanding is that Mozilla is still waiting for more data from Google about the effectiveness of our new technology, and is also considering the limited circumstances in which their users may send URLs to Google for scanning (this only happens if a page looks sufficiently suspicious). This new protection, which is designed to detect new phishing pages as well as malicious downloads, was highlighted recently on our Chromium Blog,” wrote Ian Fette, senior product manager for Chrome.

“We believe this is a reasonable solution for Chrome users, and Microsoft takes a similar approach in Internet Explorer that involves sending URLs to Microsoft. The offer remains for Mozilla to have access to our new APIs for Firefox should they choose that it’s in the best interests of their users,” he wrote.

According to that Chromium Blog post from last week, “All About Safe Browsing” Google does not hold any personally identifiable information for more than two weeks, that the data isn’t used anywhere else within Google, and that users can turn the Safe Browsing features off.

Mozilla doesn’t appear to be fully swayed — yet. “Our partnership with Google’s safe browsing team is a positive one. Their team has made phishing and malware detection services available to our users and these are already implemented in Firefox. Their new services communicate more information back to Google about a user’s browsing history, and we are still evaluating the merits of that approach,” said Johnathan Nightingale, Mozilla’s director of Firefox engineering, in a statement to CSOonline.

While Google and FireFox figure out the privacy implications, end users are left with a number of questions. The first is what level of privacy do they want to give up to improve browsing security, and secondly why — at its best — is Safe Browsing technology only 50 percent effective against these threats?

George V. Hulme writes about security and technology from his home in Minneapolis. You can also find him tweeting about those topics on Twitter at @georgevhulme.