IE8 Security Part III: SmartScreen® Filter


As someone whose email address is posted in thousands of forum posts, newsgroup discussions, and blogs, I get a lot of spam. Of the spam I receive, a significant number of messages represent phishing attacks. Most of these lures aren’t very clever or convincing, but phishing has become a simple numbers game—hosting phishing sites is cheap, and even if only a few users fall for any given phishing attack, attackers will profit by increasing the volume of phishing campaigns.

In Internet Explorer 7, we introduced the Phishing Filter, a dynamic security feature designed to warn users when they attempt to visit known-phishing sites, and worked with partners to introduce Extended Validation certificates that light up the address bar when users visit sites with verified identity information. Beyond the Phishing Filter, Microsoft has also published educational materials on identifying phishing scams, and developed a strategy to attack phishing at multiple levels.

For Internet Explorer 8, we’ve built upon the success of the Phishing Filter feature (which blocks over a million phishing attacks weekly) to develop the SmartScreen® Filter, a replacement that improves upon the Phishing Filter in a number of important ways:

  • Improved user interface
  • Faster performance
  • New heuristics & enhanced telemetry
  • Anti-Malware support
  • Improved Group Policy support

I’ll describe each of these in the sections that follow.

Improved User Interface
First, we’ve simplified the opt-in experience for the SmartScreen Filter, integrating the option into the IE first-run experience. After first-run, you can later change your preferences easily by using the option on the classic Tools menu.

Next, the bold new SmartScreen blocking page offers clear language and guidance to help you avoid known-unsafe websites. Here’s a screenshot from a recent phishing site I encountered:

SmartScreen Blocking Page

The “Go to my homepage” link enables you easily to navigate away from the unsafe website to start browsing from a trusted location. If you instead choose to ignore the SmartScreen warning by clicking the “Disregard and continue” link, the address bar remains red as a persistent warning as long as you are on the unsafe site.

If you uncover a new phishing site, you can submit it for analysis using the “Report Unsafe Website” option on the Tools menu. In the unlikely event of a false-positive, you can provide feedback using the “Report that this is not an unsafe website” link on the blocking page or by clicking the “Unsafe Website” flyout in the address bar.

Improved Performance
As a part of our overall investment in improving performance across Internet Explorer, we’ve made several performance tweaks to the SmartScreen Filter to improve its speed and lower its impact on browser performance. Detection of unsafe sites happens in parallel with navigation, so you can confidently surf the web without being forced to make a tradeoff between speed and safety.

New heuristics & telemetry
As attackers have evolved their phishing sites in an attempt to avoid being recognized and blocked, the SmartScreen Filter has also evolved to catch more phish than ever before. New heuristics, developed with help from security research teams across Microsoft, are able to evaluate more aspects of web pages to detect suspicious behavior. These new heuristics, combined with enhanced telemetry, allow the URL Reputation Service to identify and block phishing sites faster than ever.

In rare cases, SmartScreen will request feedback on sites of unknown reputation, as shown in this screenshot:

SmartScreen Feedback Request Page

User feedback about unknown sites is collected by the SmartScreen web service and quickly evaluated to block new phish as they are discovered in the wild.

Anti-Malware Support
The SmartScreen Filter goes beyond anti-phishing to help block sites that are known to distribute malware, malicious software that attempts to attack your computer or steal your personal information. There are many types of malware, but most types can impact your privacy and security. The SmartScreen anti-malware feature is URL-reputation-based, which means that it evaluates the servers hosting downloads to determine if those servers are known to distribute unsafe content. SmartScreen’s reputation-based analysis works in concert with other signature-based anti-malware technologies like the Malicious Software Removal Tool, Windows Defender, and Windows Live OneCare, in order to provide comprehensive protection against malicious software.

If you are lured to a site known to distribute malware, the SmartScreen blocking page is displayed and indicates that the server is known to distribute unsafe software:

SmartScreen Blocking Page for Server Known to Distribute Malware

On the other hand, if you click on a direct link to a download (from an instant message, for instance) hosted by a known-malicious site, the Internet Explorer download dialog will interrupt the download to warn you of the threat:

Unsafe Download Warning Dialog

SmartScreen’s anti-malware feature complemented by the IE8 features that combat malicious repurposing or exploit of browser add-ons, helps to protect you from a full range of malicious websites.

Group Policy Support
Group Policy can be used to enable or disable the SmartScreen Filter for Internet Explorer users across an entire Windows domain. A new Group Policy option is available that allows domain administrators to block users from overriding SmartScreen Filter warnings. When Group Policy restrictions are enabled, the option to override the SmartScreen warning screen is removed from the blocking pages and download dialog.

SmartScreen Warning Page with Override Removed

Privacy
As outlined in Dean’s post last week, Privacy is a core component of trustworthy browsing. As with IE7, Microsoft remains committed to helping ensure users’ privacy while providing protection from unsafe websites. URL data submitted to the SmartScreen web service for evaluation is transmitted in encrypted format over HTTPS. The data is not stored with a user’s IP address or other personally identifiable information. Because user privacy is important in all Microsoft’s products and technologies, Microsoft has taken steps to help ensure that no personally identifiable information is retained or used for purposes other than improving online safety; data will not be used to identify, contact, or provide advertising to users. You can read more in our privacy statement.

Conclusion
Web criminals are increasingly relying on social engineering attacks to engage in their criminal enterprises, but we’re working hard to deliver the tools to help keep you safe on the web. The IE8 SmartScreen Filter is designed to combat both phishing and malware sites while protecting your privacy and enabling high-performance browsing. I strongly recommend you enable the SmartScreen Filter and give it a spin in IE8 Beta 2, due in August.

Please stay tuned to the IEBlog for further posts on IE8 Security improvements!

Eric Lawrence
Program Manager
Internet Explorer Security

Comments (52)

  1. I just posted an article about Internet Explorer 8 security features . This is based on a recent briefing

  2. Kwispel says:

    What stops the phisers from using a botnet (lots of different IPs) to report their pishing sites as safe and getting around the filter?

    Is there some kind of protection against this?

  3. Jeff Parker says:

    My only question would be is it annoying? Take for example the Phising Filter in IE 7 not only is it the first thing I shut off, I am instantly reminded to shut it off when I visit the very first site in a new computer setup. Because this balloon keeps popping up and complaining.

    I am all for better security on the browser, however the Phishing filter was such an annoyance it got shut off, we even rolled shutting it off out globally in our organization because our helpdesk calls spike with users calling asking how to turn it off.

  4. AlexGl [MSFT] says:

    @Kwispel:

    We have human graders who examine reports of phishing/not phishing. A large number of reports doesn’t automatically change the rating without a person actually looking at the page in question and deciding whether it truly is phishing.

    @Jeff Parker:

    Yes, already in Beta 1, we’ve removed the annoyance factors you mention. This is part of what Eric describes as having "simplified the opt-in experience".

  5. Jamie says:

    I take it that the parallel checking will prevent the Phishing Filter problems that have been seen when using an authenticating proxy server? Phishing Filter can make the browser unusable in these sorts of setups.

  6. Techritic says:

    Can you please get your damn standards right already? I’m tired of putting half of my time trying to get my site working in Internet Explorer.

  7. Privacy Concerns says:

    All VERY good; keep it up.

    However, (I know its a bit too late in the development process) but i would love a feature, where cookies, authentication sessions, etc expire and are deleted after a number of days automatically! Like history, the user chooses how long info is kept.

    Anyone know of an addon?

  8. So from the screenshot in the “New heuristics & telemetry” section I gather that the filter will give a warning if you directly access an IP address.

    Will this warning also pop up when accessing a LAN address? E.g. 10.0.0.1 or 192.168.1.1 or 127.0.0.1? It shouldn’t, IMO, as these addresses don’t pose a phishing threat and are frequently used by developers for development purposes.

    ~Grauw

  9. Faramond says:

    Do you use mixed-script domain names as a heuristic? It seems like a warning should be triggered whenever users visit a domain name that does includes characters beyond simple ASCII and their own character set.

    You might also want to add an option to prohibit browsing of non-ASCII domains. (Non-ASCII domains are bound to lead to a big increase in phishing due to the similarity of different glyphs.)

  10. consumer4beta@hotmail.com says:

    "catch more phish"? LOL

    Btw you mention the anti-malware works in concert with Live OneCare….does this mean Live OneCare users are better protected with IE8’s SmartScreen (TM) tech? Any plans to integrate the anti-malware feature with popular anti-virus software such as Norton, Kaspersky, NOD32?

  11. I highly appreciate the functionality and aesthetics if how *this* is implemented in to IE. I also applaud emphasizing the domain name (or IP address) of the potential attack site.

    Eric, I’m surprised though that you simply don’t just use an email form to protect your email address from spammers. Unless you spend time with the Hotmail folks working on spam filters?

    PS – I see rounded corners, any chance we could at *least* get "-ie-border-radius" support in IE8? :D

  12. Internet Explorer 8 – Security

  13. Andre says:

    So the SmartScreen Filter has two buttons, Yes and No, where both will report the address to Microsoft, either as safe or unsafe.

    I’m glad I’m not using the IE anymore at all.

  14. Kwispel says:

    "We have human graders who examine reports of phishing/not phishing."

    Worldwide? Or are these Phising-lists only updated between 9h and 17h Microsoft-time?

  15. Jay says:

    I work for a bank and we get phished once every six weeks.  When I report the phish in IE, it takes too long to be included in the phishing filter.  I would expect it to take 5 minutes or less to verify and add to the filter.  Most times, I am able to shut down the site at the ISP level quicker than getting it added to the phishing filter.  The phishing filter submission is typically faster for Firefox/Google.  Is there a way you can add trusted sources/priority submissions for banks/financial institutions?

  16. EricLaw [MSFT] says:

    @Kwispel: There are grading teams evaluating reports all day, every day, worldwide.

    @Jay: Beyond user-reports, we collect phishing reports from over a dozen data providers that work with the major brand protection companies.  I’m interested in troubleshooting why you’re seeing such a long time to block; please feel free to send me a note (ericlaw at microsoft) next time you encounter a problem.  Note, however, that there’s a local cache (for performance reasons) so if you report from one machine, you should later check to see whether a block was issued from another machine.

    @John: My email address was public long before spam was a significant problem.  I made posts to newsgroups with my actual address many years ago, and I’m not inclined to switch now.  And yes, this is somewhat nice, because I get to evaluate spam and phishing filters against "real world" data.  :-)

    @someone: Think of SmartScreen’s anti-malware feature as a "first line of defense" against malware; it blocks sites known to deliver malware, which is nice because it can block even new/unknown malware distributed from sites that are known to distribute malicious software.  In contrast, OneCare AV and third-party AV tend to be signature-based– one the plus side, this means that known-malware are blocked even when distributed from unknown/new sites, but on the downside, there’s a lag between the discovery of a new piece of malware and when a signature is generated and rolled out to block that malware.  Hence, these two types of technologies work best together.

    @Faramond: Please see http://blogs.msdn.com/ie/archive/2005/12/19/505564.aspx for more information about IE’s handling of IDN names.  In that post, we describe the mitigations in place against malicious non-ASCII names.  The SmartScreen Filter will block known-malicious IDN sites, and users have the option of turning off Unicode display of IDN to completely prevent spoofing possibilities (Tools / Internet Options / Advanced / Always show encoded address).

    @Laurens Holst: When evaluating a site, the fact that an IP-address was used rather than a hostname is only one factor used in the evaluation, for the reason you describe– in many cases, navigation to IP-only sites is an innocuous daily task for IT-professionals and developers.

    @Jamie: I’d be very interested to learn more about the problems you’ve encountered with authenticating proxies.  Please send me a note (ericlaw at microsoft) with more info.

    Thanks, all!

  17. Mirronelli says:

    Really nice and clear.

    One suggestion: If the possibility to continue and disregard the warning is disabled by administrators the smartscreen filter should state this clearly and not just tell that you only can go to homepage. Users will blame the browser or windows for this and not their own administrators.

    To Andre: Where did you come up with the thing that both buttons (red and green) will send report to MS? Actually neither of them will. You must click the link: "Report this site …" to send a report.

  18. IE8 Security Part III: SmartScreen® Filter

  19. Rocky says:

    Dear Eric:

    Is the anti-malware or anti-phising provider is open or only can supplied by Microsoft? Like in firefox, people can use both firefox’s own database or Google’s database. I think if the provider is open, maybe many professional security company could supply their solution for anti-phising and anti-malware, maybe it’s a good thing for the end-users :)

  20. CableGuy says:

    Why not use blacklists as Firefox does?

  21. TechBlog says:

    [Note: Techblogger Claus Valca wrote an excellent guest post on June 29 about issues surrounding the popular AVG Free antivirus program. Since then, AVG has taken steps to fix problems with its new LinkScanner feature, and Claus has been kind…

  22. Nektar says:

    The problem with IE7’s Phishing Filter is that it is off by default and most users never care or even know to turn it on. "Phishing Filter!" they say, "What is that? Let’s turn it off."

    Users do not bather to change the defaults. Even educated users that I know of, do not care to check that the Filter is on or even care for its presence. Users are busy. They have more important things to do than have to configure IE options. They want it to work as best as possible out of the box. The same with the default search provider. Most users, almost all users, do not care to click "Change my default provider" and then to navigate and scroll through a page-full of providers. Please improve this experience (A) by integrating the Phishing Filter and search provider preferences into the Set-up process instead of the first-run experience so that users will give it more attention, (B) by simplifying the search provider choice dialog, putting a list-box of search provider choices in front of users during IE8 set-up instead of presenting a "confusing" full Web-page of search engines with descriptions at first-run and (C) by turning the Phishing Filter on by default. What do you think?

  23. EricLaw [MSFT] says:

    @Nektar: Actually, a *significant majority* of IE7 users do turn on the Phishing Filter.  Remember, there are a number of prompts during initial use, and if the user configured the "Use recommended settings" during Vista setup, the filter is on by default for them.

    Integrating these choices into setup rather than first run wouldn’t really work because  only one user on a computer runs IE setup, but other users of the same computer may have different preferences.  Because First-Run is per-user, the current design provides the opportunity to set their defaults as desired.  As you’ll see in Beta-2, we’ve significantly streamlined the first run experience.  

  24. phish-shield says:

    please do not develop the solutions "made for geeks, made by geeks". Look at the simple approach of phishing-shied from everyday user’s point of view, and develop something simple as presented by http://ww.parentapproval.com/

  25. Most of the phishing solutions are not transparent at user-level, and seems like "made for geeks, made by geeks".

    Look at the simple approach of phishing-shied from everyday user’s point of view, and develop something simple as presented by http://www.parentapproval.com/

  26. Ted says:

    @@phish-shield:

    Hmm… IE8 shows a big red blocking page that says "This is a phishing site.  STOP!"  That doesn’t really seem like it lacks "transparency", vs the "parent-approval" toolbar, which involves multiple configuration UI, allow lists, dozens of checkboxes, and the requirement that every website be entered manually for "allow" or "deny."  

    Couple that with the ludicrous "patent pending" claimed by the "parent approval" company, and you can bet that Microsoft isn’t going to implement something like that.  

    Methinks maybe you work for those patent trolls and are hoping microsoft will do something that gets them sued?

  27. a {color : #0033CC;} a:link {color: #0033CC;} a:visited.local {color: #0033CC;} a:visited {color : #800080;}

  28. Igor Macori says:

    Si sta avvicinando a grandi passi il rilascio della Beta 2 della versione 8 di Internet Explorer . Come

  29. The next beta for Internet Explorer has been released for broad distribution to the public, according

  30. IEBlog says:

    Hello, My name is Sébastien Zimmermann. I’m the developer owner for the Visual Search Feature , which

  31. Привет, меня зовут Себастьян Циммерман (Sébastien Zimmermann) и я являюсь основным разработчиком функции

  32. Привет, меня зовут Себастьян Циммерман (Sébastien Zimmermann) и я являюсь основным разработчиком функции

  33. IEBlog says:

    Back in June, Dean Hachamovitch kicked off a series of blog posts explaining how the IE team approached

  34. IEBlog says:

    Hello, I’m Alex Glover and I’m the test owner of the SmartScreen Filter in Internet Explorer 8. The SmartScreen

  35. Изменения в фильтре SmartScreen в IE8 RC1 Привет, меня зовут Алекс Гловер (Alex Glover) и я являюсь главным

  36.     이메일 주소를 포함한 글이 포럼이나, 뉴스그룹, 블로그 등에 올라가게 되면 엄청나게 많은 양의 스팸을 받게 됩니다. 그렇게 받는 스팸 중에는 피싱 메일이 상당수를

  37. A study by NSS Labs of 6 major web browsers shows a large difference in their ability to block "socially

  38. After my post about IE8 last week I got an email from someone asking me to explain more about other security

  39. Maria's blog says:

    After my post about IE8 last week I got an email from someone asking me to explain more about other security

  40. Maria's Blog says:

    After my post about IE8 last week I got an email from someone asking me to explain more about other security

  41. After my post about IE8 last week I got an email from someone asking me to explain more about other security

  42. Maria's blog says:

    After my post about IE8 last week I got an email from someone asking me to explain more about other security

  43. Maria's blog says:

    After my post about IE8 last week I got an email from someone asking me to explain more about other security

  44. After my post about IE8 last week I got an email from someone asking me to explain more about other security

  45. My colleague in New York, Peter Laudati , just alerted me to this report in which IE8 was found to be

  46. IEBlog says:

    Over the last year, we’ve published two posts about how the IE8 SmartScreen ® filter helps to prevent

  47. Безопасность IE8: защита от вредоносного ПО с помощью фильтра SmartScreen В прошлом году мы опубликовали

  48. I attended Scott Charney’s keynote this morning at RSA – Moving Towards End to End Trust: A Collaborative

  49. Aus aktuellem Anlass möchte ich doch einmal darauf eingehen, in wieweit aktuelle Browser den Nutzer vor potentiell gefährlichen Seiten schützt. Wichtig ist in diesem Zusammenhang, dass ich mich bei diesem Beitrag jeweils auf die aktuellste Generation

  50.     안녕하세요. 이번에는 Jon DeVaan 이 최근에 UAC 에 대해 받은 피드백에 대해 이야기하겠습니다.  Windows 7 을 완성하기 위한 작업의

  51. Nino Iaccarino says:

    Two of our our legitimate websites have been blocked by this "Filter" This is unacceptable, that a disgruntled user or user(s) can affect the way that a website is presented to other end users.

    The two website I am referring to are:

    http://www.endeavour.edu.au

    http://learn.endeavour.edu.au

    We are an educational institution and in no way are our sites associated, affiliated or have any phishing or malware embedded into our site for end users.

    Not providing companies a way to even view or amend these incorrect false positives is appalling. Also, the fact that real end users at Microsoft review these sites is crazy.

    Prove to me that our sites are malicious, send me a report, contact me, anything…..and I will gladly take the sites down until the issues have been resolved.