Handler on Duty: Jan Kopriva
Threat Level: green
Podcast Detail
SANS Stormcast Thursday, August 27th, 2026: Entra ID Admins; Unifi Patches; log4j Vuln; Sleepwalker Malware
If you are not able to play the podcast using the player below: Use this direct link to the audio file: https://traffic.libsyn.com/securitypodcast/10070.mp3
My Next Class
Click HERE to learn more about classes Johannes is teaching for SANS
Who Has Admin Rights in your Entra ID Directory?
https://isc.sans.edu/diary/Who%20Has%20Admin%20Rights%20in%20your%20Entra%20ID%20Directory%3F/33284
Ubiquity Unifi Patches
https://community.ui.com/releases/Security-Advisory-Bulletin-067/fc4a3488-7c43-4628-8bab-f715e96dbfc9
Log4J FilteredObjectInputStream Vulnerability
https://github.com/joanbono/log4j2-4255-exploit
https://jeffmcjunkin.com/posts/log4j2-fois-marshalledobject/
Sleepwalker Malware
https://r136a1.dev/2026/08/24/sleepwalker-a-passive-backdoor-with-its-own-command-language/
My Upcoming Classes
https://www.sans.org/profiles/dr-johannes-ullrich
| Application Security: Securing Web Applications, APIs, and Microservices | Las Vegas | Sep 21st - Sep 25th 2026 |
| Network Monitoring and Threat Detection In-Depth | Amsterdam | Oct 12th - Oct 17th 2026 |
| Application Security: Securing Web Applications, APIs, and Microservices | Washington | Dec 14th - Dec 18th 2026 |
| Application Security: Securing Web Applications, APIs, and Microservices | Online | India Standard Time | Mar 15th - Mar 19th 2027 |
Podcast Transcript
Hello and welcome to the Thursday, August 27th, 2026 edition of the SANS Internet Storm Center's Stormcast. My name is Johannes Ullrich, recording today from Jacksonville, Florida. And this episode is brought to you by the SANS.edu Graduate Certificate Program in Cybersecurity Engineering. Well, Rob is still at it, creating useful PowerShell scripts to inspect Entera ID configurations. The latest one looks at who has administrator rights. Now, this is a little bit more complicated than just looking for like administrators because, well, there are a bunch of different roles with some level of administrator privileges. There's not just the global administrator. There's, for example, also something like with global read permissions where, well, a user that basically can read everything, which in itself may not always be as defined as administrator, but certainly it's a role to keep an eye on. And what Rob's PowerShell script here does is basically collect all of the different roles and then either display how many users you have for each one of these roles or who the actual user is associated with this role. So quick little audit script that's certainly quite useful to find like leftover administrators that are no longer needed or maybe something malicious that was added by NetHacker. Ubiquiti today released Security Advisory Bulletin number 67. This bulletin does outline 22 different vulnerabilities that affect several products sold under the UniFi brand. Now, when you're patching these devices, keep in mind that, for example, if you have one of their dream machines or such, you may have to patch the operating system itself and then also applications that you installed on the device, like, for example, the Protect or the Network or the Talk product that is being sold under the UniFi brand. The vulnerabilities themselves, some of them have rather high CVSS scores, like 10 and there are a few 9.9s and 9.1 vulnerabilities there. Some of the CVSS scores look kind of high, like there is a approach escalation vulnerability with a score of 9.9. However, the description is extremely sparse, so really hard to second guess here the rating for the CVSS scores. Also, a number of these vulnerabilities were apparently found by AI tools, which these days, of course, is no big surprise. As far as mitigations go, the bulletin doesn't really outline any, but pretty much all of the vulnerabilities from me scanning them do require access to the control plane, basically access to the respective UniFi application. So, just like with all similar applications, try to limit who can access your network configuration, your camera configuration systems and the like. And that will significantly reduce the exposure and with that the risk of exploitation. The actual patches were released over the last few days, so if you have auto update enabled, you're probably already set. And then we do have a new log4j vulnerability. Now, as this vulnerability became known, of course, there was a lot of talk about yet another log4j disaster. This one doesn't appear to be that severe, that bad, I should say. It's certainly a remote code execution vulnerability. It's a deserialization flaw in the filtered object input stream feature. The problem with this function is it's actually meant to do the right thing. You have deserialization because you do allow arbitrary objects to be deserialized. Well, the filtered part in filtered object input stream is supposed to exactly prevent this. But as so often, well, they're allowing one particular object in here that then actually can in turn itself create arbitrary object. So that gives you that famous gadget and with that remote code execution. One of the big constraints here is to for this to be exploitable is that you must have software that directly accepts these log4j serialized objects. That's typically not the case. So this is not like the old log4j vulnerability where I could send a request to a web server. And as that request is being logged, it would trigger the vulnerability. As a result, this is probably not exploitable in most software that uses log4j. So, I'll also link to Jeff's blog post in addition to the proof of concept exploit. So, yes, an exploit is known, is public. No patches available at this point and not even a CVE number yet. Apparently, the bug report that this was originally reported with wasn't made like hidden as usual security reports, but was left public for a couple of days. And that's sort of how this entire issue became public. And we got some interesting malware analysis from Dominic with Palo Alto. So, this malware is described as sleepwalker. And what makes it kind of unique and sort of also caught my attention is that it has a little bit unusual, at least these days, command control system. Usually these days you have the malware reach out to some kind of web server or such in order to retrieve instructions. Well, back in the old days, I've seen quite often sort of some passive listening for packets. There were some famous DDoS agents and such that did it. Well, this is exactly what this particular malware does. Also interesting that it sideloads itself into the antivirus product, ESET. And well, then it starts listening on the network for packets that match very specific characteristics. It actually doesn't appear to matter what protocol, what port, but then the payload has to match specific checksum conditions. If that's the case, they're then decrypted and interpreted as bytecode, which then instructs the malware to do various additional operations. So, interesting malware here and actually also a real nice and interesting write-up of some rather unusual malware that probably often is being missed because it doesn't have sort of that today classic kind of beaconing and such a behavior that you would typically see in malware. Well, and this is it for today. So, thanks for listening. Thanks for liking. Thanks for subscribing. And talk to you again tomorrow. Bye.





