In partnership with

~6 MIN READ
FACTAbout 70% of the countries Russia's Turla has targeted since 2022 are former Soviet republics or European countries with historical Soviet ties. (ESET via The Record, November 2026)
The Signal
 
This week's break-ins came through the side doors: add-ons, hijacked developer accounts and a government mail server that doubled as a spy's base. Here's what got opened, who's behind it and what to patch before the weekend.

PS: Was this forwarded to you? Subscribe free at exzeccyber.com/subscribe →

In this edition
 📌 Big Cyber News
 🚨 Can't Miss
 🤖 AI in Cyber
 🕵️ Threat Intel
 🛠️ Tools & Tactics
 🧪 Strange Cyber
📌 Big Cyber News
 
MSP / Supply Chain
288 Computers in 90 Minutes, All Through One ScreenConnect Add-On
Intro
Nobody had to break ScreenConnect itself this time. The intruder came in through an add-on bolted onto it.
What Happened
Huntress says someone used several cloud-hosted ScreenConnect panels to push rogue copies of the remote access tool onto 288 computers at 48 organizations within about 90 minutes. It traced the activity to RMM+, an independent extension that skips ScreenConnect's accounts, roles and MFA for a shared access key, blank by default in two versions. The toolkit, SCKit, also tried replaying stolen session cookies against other customers' panels. ConnectWise disabled RMM+ and four other extensions in about 400 partner environments; on-premises customers have to do it themselves.
Why It's Important
One remote management panel reaches every client an MSP manages, so an add-on that sidesteps MFA works like a skeleton key. Blackpoint says the kit also exploits CVE-2026-84869, September's 9.9 ScreenConnect flaw, which hits clients that were never patched.
The Other Side
ConnectWise says it found no compromise of its infrastructure and no flaw in ScreenConnect itself, and the cookie replays failed in every log Huntress checked. How the intruder first reached the panels is still unknown, and nobody has been named.
 
👉 Takeaway
If you or your MSP runs ScreenConnect, check for RMM+ today, and don't treat switching it off as cleanup. ConnectWise warns that alone "may not remove unauthorized software, persistence mechanisms, or access."
TL;DR: A third-party ScreenConnect add-on let an intruder plant rogue remote access at 48 organizations.
Further reading: The Record · CyberScoop
🚨 Can't Miss
 
 
Supply Chain
Starting Tuesday, an attacker opened test pull requests in 61,764 repositories, then slipped a stealer into ones the stolen accounts could write to. Aikido says it ran in at least 1,806, grabbing Actions secrets, cloud keys and package tokens, then reverted changes and deleted logs. How the accounts were stolen is unknown.
→ Search your org for a wfx-docs branch, and rotate every secret an affected repo could read.
 
Mobile Exploits
iVerify says the leaked DarkSword and Coruna exploit kits are now crimeware. One traffic seller bought an expired analytics domain that sites still loaded and used it to fire DarkSword at Safari on iOS 18.4.0 through 18.7.2. Credential and crypto theft made up three-quarters of detections, and every enterprise victim was at or below managing-director level.
→ Phones still sitting on iOS 18 are the target. Update them.
 
Takedown
Europol and nine countries took down 235 servers and 257 domains behind three malware kits that feed ransomware crews, and found credentials tied to 6.4 million victims. Gootloader, which hides in fake legal documents on hacked WordPress sites, lost 218 domains, and 681 infected sites were cleaned. No arrests were announced.
→ Check Have I Been Pwned, and put MFA on your WordPress admins.

Are your sales foundations ready to scale?

You’re growing from 10 reps to 100. Your agents can only work with the data and processes you give them, and the foundations you set now decide whether your systems hold.

Attio’s agentic sales audit looks at where you are today and shows you where to start, whether that’s your data foundation or a fuller agentic sales process.

🤖 AI in Cyber
 
 
AI Attacks
Malbear Labs says a Chinese-speaking crew used an autonomous framework built on a jailbroken DeepSeek Flash model to enter a European telecom through CVE-2026-20320, a Cisco BroadWorks flaw fixed in August. Over two weeks it worked through 13 companies hosted there, taking dozens of credential sets and thousands of rows of subscriber billing records.
→ Patch BroadWorks. The AI only needed a bug you could have fixed in August.
 
AI Agents
United will block AI agents from managing reservations and buying tickets, on its own platforms and through third-party agencies. It says handing your login to an AI assistant "can expose sensitive information" and can trigger unwanted purchases; check-in needs a human to confirm nothing banned is in the bag. It hasn't said how it will spot the bots.
→ Expect more companies to tell customers: don't give an agent your password.
🕵️ Threat Intel
 
 
Ransomware
Qilin claims 365-plus victims in the past three months and gives affiliates at least 80% of each ransom. Krebs followed its admin, "Haise," through old forum handles to an email that once used the password "Rjun132," then to a leaked Russian government-services record. He says the trail points to Aleksey Yakovtsev, 36, of St. Petersburg. Yakovtsev didn't respond; police haven't confirmed it.
→ Reused passwords outlive everyone's opsec, even a ransomware boss's.
 
Espionage
ESET found Turla's new Cereal toolkit on Kazakh and Romanian government Exchange servers; in Romania it was set to grab email to and from NATO and EU addresses, picking out Word and PDF attachments. There was no outside command server; operators talked to the hacked Exchange server itself. How they got in is unknown.
→ On Exchange, watch what runs inside the mail transport service, not just outbound traffic.
🛠️ Tools & Tactics
 
 
Patch Now
CVE-2026-119618 is a race condition in Apache's HTTP/2 module that Red Hat rates 9.8 and that could let an unauthenticated attacker run commands. Anthropic and Google researchers found it separately; versions 2.4.55 through 2.4.70 are affected. No attacks reported yet, but a public proof of concept crashes servers. The play: update to 2.4.71 or your vendor's package; if you can't, disable HTTP/2, since HTTP/1.1-only servers aren't affected. A WordPress update won't fix this one.

See what's really happening in every deal

Aligned shows you what buyers are doing between meetings inside your most important deals then surfaces the risks, openings, and next steps. No more guessing where a deal stands or finding out a champion went quiet too late. Just real visibility, so you can act before it's too late.

🧪 Strange Cyber
 
Strange but real
A Programmer Used a Broken Certificate Authority to Swap His TV's Ads for Family Photos
Intro
Most people mute the ads on their smart TV. Daniel Vrátil replaced them with pictures of his family, using a certificate authority's busted keys.
What Happened
Indian CA eMudhra, which says it issued more than 60 million certificates, had eight intermediate keys that turned out to be weak; one of them was the number four. Vrátil, a Czech developer, fed those keys to mitmproxy, minted certificates for Samsung's ad servers and pointed his 2022 Samsung TV at a box on his home network. Ad requests now get family photos from his NAS, and tracking calls get dropped. It works because Samsung hadn't shipped firmware that stops trusting the bad certificate. He posted it on GitHub as samsung-tv-adspoofer.
Why It's Important
A device that still trusts a broken CA will believe anyone holding its keys. Vrátil says the same trick could serve "countless much more sinister applications."
The Other Side
In his words it's "not exactly a one-click setup," and a Samsung update that revokes the certificate ends it. eMudhra has left the certificate business and blamed a buggy key-generation script.
 
👉 Takeaway
Revocation only works when devices get the update, so patch the TV too, even if the ads come back.
TL;DR: A developer used eMudhra's broken CA keys to swap his Samsung TV's ads for family photos.
Further reading: The Register

Watch on demand: Tabs + PwC break down how finance teams are operationalizing usage-based pricing — without the manual overhead.