Microsoft Corp. today outlined new security features that it plans to add to Internet Explorer (IE) next month, including anti-malware protection to match tools similar to those offered by its rivals and a filter the company said would block most cross-site scripting attacks.
Internet Explorer 8 Beta 2, which Microsoft has slated for release sometime in August, will include two new security tools, said Austin Wilson, the director of Windows client product management.
One, dubbed "SmartScreen Filter" by Microsoft, adds malware blocking to the antiphishing protection already embedded in IE7. The new feature, which will resemble the defenses already used by rival browsers Firefox 3.0 and Opera 9.5, will warn users when they're about to visit a site known or suspected of spreading malicious code and then block any download from that site.
Unlike Mozilla Corp.'s Firefox, which retrieves a blacklist several times daily, then stores it locally to compare against Web site addresses, IE8 will dynamically determine whether a site is potentially dangerous by pinging remote servers each time a user tries to reach a page.
Microsoft will use multiple third-party sources to compose the blacklists for both phishing and malware-hosting sites, said Wilson. It will also draw on data gathered by Windows Defender, the company's free antispyware tool. Wilson would not disclose the third-party information providers, however.
"We get the data feeds and update our lists multiple times a day," he said. "And IE8 makes the call to the URL reputation service servers, and if it's a phishing or malware site, the browser navigates away from the page and displays a warning."
He denied that the process would have a noticeable effect on IE8's performance. "Our choice was to make sure that the user has the most recent data possible," he said. "We do an asynchronous call, so the page rendering takes place while the call is made to the reputation servers."
Also to debut next month in IE8 Beta 2 is an integrated filter that Microsoft said would prevent most cross-site scripting attacks. "Today, the end user can be doing all the right things, checking the URL to make sure it's legitimate, only going to trusted sites, but because of vulnerabilities on the Web server side, they can still be compromised," said Wilson, referring to cross-site scripting attacks, which are most commonly used by identity thieves and have been on the upswing.