-
Notifications
You must be signed in to change notification settings - Fork 88
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add entry for - Mobile Security Misconfiguration > Copy/Paste Sensitive Data to Global Clipboard #150
Comments
I believe this should be included, especially given the prevalence of malicious applications on Android and the relative direct means to prevent this on a per-field basis. Example mitigation on
|
We are definitely lacking mobile entries in the VRT and I think this would be a good addition. +1 for a P4. |
While this is a good idea to add as an entry in VRT, I think it would be better to separate this in two. So there might be cases where the input content might be sensitive as mentioned (SSN, CC or similar), and there might be cases with non-sensitive content (everyday non-relevant information). |
Wasn't the concern about having a P5 variant that clients may consider fixing them? I don't think non-sensitive copy functionality is even worth Informational. |
If there are non-sensitive information then I do not see any impact or any attack scenario there, so why pushing the client to fix such cases. I think this is what we were following until now in VRT with other entries as well. |
I tweaked the names a little bit @trimkadriu but it's the same idea. P5 if not an issue and P4 if the content is sensitive. |
I have the same confusion as @trimkadriu, if its admittedly not a security issue (in the case of non-sensitive content), why put it in the VRT at all? |
There's multiple benefits of doing so e.g.:
We've been using this binary variant concept across the VRT with good results. P5 does not automatically mean Informational, in most cases it is a won't fix. In fact there are multiple N/A's that are classified as P5 for the same reasons as above. |
Having it in there explicitly helps ASEs for when a researcher will invariably submit a non-sensitive form, when we have something to push back with that is concrete it helps reduce friction. |
This should be at most a P5 for sensitive content, but I believe it should be removed from the paid program altogether. This control is a reasonable defense in depth for some very sensitive use cases that do not care about usability. It does not represent an actual vulnerability unlike many of the other P4s. It can only be taken advantage of with user interaction in conjunction with malware being present. Malware is more prevalent on PCs and macs than it is on mobile devices and disabling copying/pasting is not a common recommendation for web sites. Web security is far more established than mobile security so we should take some hints from the community there. Users use the clipboard for legitimate purposes and expect it to work. Adding these restrictions can make apps less usable for real people. There are situations where people need to copy personal/sensitive data between apps or within different views of the same app. There are even use cases like password managers where disabling clipboard usage on sensitive inputs will harm security overall because users will pick weaker passwords when they realize they can't user their copied password. This was rejected from OWASP's mobile recommendations for these reasons: OWASP/owasp-masvs#117 |
@devinlundberg I thought I'd make a comment to tag you and point out that a lot of discussion is happening in #256 this past week regarding this. You may want to chime in. |
We currently don't have an entry for this vuln type. The idea is that if a mobile app allows copy/paste functionality on sensitive inputs (such as CC/SSN) a user could use the copy functionality which would leave the data exposed in the clipboard. Any rogue app on the victim's device can read the clipboard.
The barrier for exploitation is pretty high but there is a simple remediation for this on a per input basis. Initial general consensus for this seems to be P4.
The text was updated successfully, but these errors were encountered: