05:39 PM
Gadi Evron
Gadi Evron
Connect Directly

Mozilla's Add-On Policies And Spyware Surprises

I've been using FlashGot on and off for years. It is a useful plug-in that helps you download multiple files from the same Web page "automagically." So when Firefox informed me about a new update for an add-on I've used for years, I clicked "OK" and updated it, only to find a surprise the next time I used Google.

I've been using FlashGot on and off for years. It is a useful plug-in that helps you download multiple files from the same Web page "automagically." So when Firefox informed me about a new update for an add-on I've used for years, I clicked "OK" and updated it, only to find a surprise the next time I used Google.The surprise: On top of my search results were "search refinement" suggestions in the regular Google font. Because it looked very much like Google's "did you mean...?" typo correction feature, I thought it was a new Google feature. I clicked on a suggestion and found myself surfing a different search site. To make sure I didn't just encounter malware or some form of hijacking, I clicked the Back button and examined the Google search results more closely.

Next to these recommendations was smaller text that stated they were a feature of FlashGot. While I appreciate the information, it doesn't absolve it, and I immediately uninstalled the application.

Why, you ask?

1. It installed a new substantive feature that is also considerably outside the scope of what I signed up for, without asking for my permission.

2. It changed the content of pages I was viewing without fair warning.

3. It then led me to a different Web page altogether for my searches, making this new change an advertisement and, thus, adware.

4. I suspected, although I never checked (so this is only unverified speculation), that since FlashGot is providing me with "search refinements," it might also be sending my searches to a third party, which, if true, would make it a spyware.

5. Because it never asked for me to opt-in, which is the way spammers operate, in my mind it was now a spammer. It just allowed me to disable it if I wanted to dig into the FlashGot configuration.

6. It broke my trust. What else might it be doing on my computer?

I sent an email to the funsec mailing list, telling the security/privacy/hacking/fun community that this Mozilla add-on was now spyware (see below). Immediately, a friendly guy told me of an add-on he used to open PDF files, PDF Downloader, which without warning started redirecting him to a commercial service. The knowledge this happens more commonly didn't make me feel any better.

Some of the folks on the list work at Mozilla's security department, and they were extremely helpful. I haven't seen many vendors that actively interact with the community, and whenever this happens I am impressed, both by them and by watching their reputation and respectability increase.

First, we made sure it was really FlashGot by the company called InformAction because there is an unrelated application called FlashGet. Then I made sure our definitions were the same because what I define as spyware could be considered legitimate by others, and I didn't want to file a false report.

The Mozilla security guys confirmed my report and said it is not in their purview, but they would happily relay it to the relevant team within Mozilla for further examination to see if this change/addition violates Mozilla's policies.

Apparently, the new feature added the "search refinements" to Bing and Yahoo, as well, and sends users to a site called "Surf Canyon."

Mozilla's policies aside, I wonder if Google, Microsoft, and Yahoo cover alteration of their Websites remotely for third-party advertising purposes in their AUPs, especially given they are in the online advertising business. I am not sure that would fall under Fair Use or be enforceable, but it is certainly an interesting question.

I wrote back:

If it isn't [taken care of], we can always shame FlashGot.

This may not be covered by current [Mozilla] policies, but as we have seen time and time again, legalities often come following new technologies rather than legal systems expecting them. And when abuse policies are tough, offenders find ways around them.

By letter of the law or not, this *feels* wrong. So I am hopeful Mozilla will do something about it. However, I can't really blame them if they can't.

I am unsure that an AUP *anywhere* currently covers that "apps" can provide only with features users agree to, or that they should need to notify of a major change in functionality.

It's certainly a very interesting question.

The good old comp.virus FAQ defines a Trojan horse as functionality which if the user knew what it did, he or she wouldn't be happy about it. In reverse, this fits quite well.

Let's see what happens.

Thank you very much for taking a look at this.

FlashGot's most recent update can no longer be found on the Website, and it is my understanding that Mozilla is in communication with FlashGot's creators to resolve this matter and enable FlashGot to work with Mozilla's policies. I'd feel badly for FlashGot potentially losing its advertising contract, only they didn't quite bother with my feelings or being fair, did they now?

I heard back from yet another friendly Mozilla security guy, who introduced me to the interesting "No Surprises" clause:

Changes to default home page and search preferences, as well as settings of other installed add-ons, must be related to the core functionality of the add-on.

Mozilla's policy also clearly states that:

* The add-on description must clearly state what changes the add-on makes. * All changes must be 'opt-in', meaning the user must take non-default action to enact the change. * The opt-in dialog must clearly state the name of the add-on requesting the change. * Uninstalling the add-on restores the user's original settings if they were changed.
I find the No Surprises clause fair and surprisingly elegant. While plug-ins, add-ons, and applications apps have been around for a long time, their popularity has increased dramatically these past two years due to the introduction of Facebook and iPhone applications.

While I recognize that many software products act in some ways I wouldn't approve of -- some even malicious -- I can't help but wonder what else I may be purposefully running on my computer without knowing it, and what other trickswill appear in the future.

With viruses and other types of malware, I can get upset, but I recognize it is criminal activity. Legitimate products acting abusively are what catch me off-guard.

But it is commonplace, and the amount of control we have over our machines decreases daily. This is why I am most happy to see Mozilla ahead of the curve it and that it even incorporates it into its policies. I hope they will consider some closer monitoring of already approved add-ons in the future, although I don't imagine that will be easy.

I would especially like to thank Reed Loden and Daniel Veditz at Mozilla's security team for their timely responses, openness, and professionalism.

Follow Gadi Evron on Twitter:

Gadi Evron is an independent security strategist based in Israel. Special to Dark Reading.

Comment  | 
Print  | 
More Insights
Register for Dark Reading Newsletters
White Papers
Flash Poll
Current Issue
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
Published: 2014-08-19
The SNMP module in Cisco NX-OS 7.0(3)N1(1) and earlier on Nexus 5000 and 6000 devices provides different error messages for invalid requests depending on whether the VLAN ID exists, which allows remote attackers to enumerate VLANs via a series of requests, aka Bug ID CSCup85616.

Published: 2014-08-19
The EJB invocation handler implementation in Red Hat JBossWS, as used in JBoss Enterprise Application Platform (EAP) 6.2.0 and 6.3.0, does not properly enforce the method level restrictions for outbound messages, which allows remote authenticated users to access otherwise restricted JAX-WS handlers ...

Published: 2014-08-19
The isCallerInRole function in SimpleSecurityManager in JBoss Application Server (AS) 7, as used in Red Hat JBoss Enterprise Application Platform (JBEAP) 6.3.0, does not properly check caller roles, which allows remote authenticated users to bypass access restrictions via unspecified vectors.

Published: 2014-08-19
RESTEasy 2.3.1 before 2.3.8.SP2 and 3.x before 3.0.9, as used in Red Hat JBoss Enterprise Application Platform (EAP) 6.3.0, does not disable external entities when the resteasy.document.expand.entity.references parameter is set to false, which allows remote attackers to read arbitrary files and have...

Published: 2014-08-19
The (1) serf_ssl_cert_issuer, (2) serf_ssl_cert_subject, and (3) serf_ssl_cert_certificate functions in Serf 0.2.0 through 1.3.x before 1.3.7 does not properly handle a NUL byte in a domain name in the subject's Common Name (CN) field of an X.509 certificate, which allows man-in-the-middle attackers...

Best of the Web
Dark Reading Radio
Archived Dark Reading Radio
Dark Reading continuing coverage of the Black Hat 2014 conference brings interviews and commentary to Dark Reading listeners.