Hi all,
If you're just now signing in for the first time in 12+ hours, you may just now be finding out that Lemmy World and other instances where hijacked. The hijackers had the full abilities of hijacked user, mod, and admin accounts. At this time, I am only aware of instance defacing and URL redirections to have been done by the hijackers.
If you were not forced to sign back in this morning, contact your instance admin to verify mitigations were completed on your instance.
How?
This occurred due to an XSS attack in the recently added custom emojis. Instance admins should follow the issue tracker on the LemmyNet GitHub, as well as the Matrix Chat. Post-Incident Activity is still on-going.
Currently, it is likely that just your session cookie was stolen, with instance admins being targeted specifically by checking for navAdmin, an HTML element only instance admins had. I do not believe this to affect users across instances, but I have yet to confirm this.
What happens next?
As I am not the developers or affected instance admins, I cannot make any guarantees. However, here is what you'll likely see:
- Post Incident investigation continues. This will include inspecting code, posts, websites, and more used by the hijackers. An official incident writeup may occur. You should expect the following from that report:
- Exactly what happened, when.
- The incident response that occurred from instance admins
- Information that might have helped resolve the issue sooner
- Any issues that prevented successful resolution
- What should have been done differently by admins
- What should be improved by developers
- What can be used to identify the next attack
- What tools are needed to identify that information
-
A CVE is created. This is an official alert of the issue, and notifies security experts (and enthusiasts), even those not using lemmy, about the issue.
-
A code security audit is done. This will likely just be casual reviews by technical lemmy users. However, I will be reaching out to the Mozilla Foundation and Cure53 as they recently did an audit of Mastodon. If there is interest in an external audit of lemmy and the costs are affordable, I'll look into crowdfunding this cost.
XSS, classic
Thanks for your hard work, fuck the trolls that always have to poke holes in shit
Honestly, I see it as a win.
The people that did this didn't really act out in a coordinated attack. They were just kind of playing around, redirecting to lemonparty, changing page elements.
It could have been a lot worse. The site could have been redirecting to malicious websites, downloading trojans, doing a lot of bad things. Instead, we got direct attention to the security vulnerabilities in question, and they're being worked on and patched out relatively quickly. Helps that a lot of those on these communities are focused in programming and cybersecurity.
The fundamental issue is not that emoji XSS (that's just a vector), but how JWTs are implemented and [not] secured. I've read that it was reported at least this January (https://akkoma.nrd.li/notice/AXXhAVF7N5ZH1V972W).
So, developers were already aware, yet - as I'm checking 0.18.1 - they have not fixed the
unsafe-inlineandunsafe-evalCSP, haven't madejwtcookie HttpOnly, and haven't done anything aboutexpandjtiin the JWTs. I hope the recent events will make them do to so, and not just patch this particular XSS.Kegan may have been better off submitting a GitHub issue. And if issues haven't been open for these critical security flaws, you should, too.
That's fair, I was a little harsh. Not all hackers are looking to cause damage!
Agree completely. In the grand scheme of things the damage that appears to have happened here is small potatoes, but it brought attention to the vulnerability so it was patched quickly. Going forward now, the authors and contributors to the project might be a bit more focused on hardening the software against these types of vulnerabilities. Pen testing is invaluable on wide user-base internet accessible platforms like this because it makes better, more secure software. Unfortunately this breech wasn't under the "ethical pen testing" umbrella but it sure as hell brought the vulnerability to the mindshare of everyone with a stake in it, so I view it as a net win.
Yes, exactly. This could've been way worse.
thanks for the update