Watch the 2-minute recap
The US says it has disrupted a North Korean scam that stole more than $400 million in crypto using fake video-meeting invites. Reports also covered a Cursor breach and an exploited chat-server flaw. The advice is simple: use official apps, confirm unusual requests, and update promptly.
Today at a glance
- US disrupts North Korean FaceStealer crypto scam. [BleepingComputer]
- Cursor breach puts a fake Claude Code extension in its marketplace. [BleepingComputer]
- CISA says attackers exploit an UnrealIRCd flaw. [The Hacker News]
- CrowdStrike report: attackers use AI for personalised phishing. [CyberDWN feed]
- Salesforce guide: AI guardrails need human oversight. [CyberDWN feed]
- FAA ground stop shows no sign of a cyberattack. [SecurityWeek]
- Apple explains Secure Exclaves for camera and microphone data. [SecurityWeek]
- Microsoft releases emergency Outlook patches. [CyberDWN feed]
US disrupts North Korean FaceStealer crypto-theft campaign
The US government says it has disrupted a North Korean social engineering campaign called FaceStealer, which BleepingComputer reports stole more than $400 million in cryptocurrency. Victims were reportedly sent fake video-meeting invites and told to run a “fix” after a fake error. As advice, join meetings only through the official app, and never run a script or installer offered as a fix. [BleepingComputer]

Cursor breach puts a malicious Claude Code extension in its marketplace
Cursor has confirmed a breach. BleepingComputer reports that hackers distributed a malicious Claude Code extension through its marketplace after stealing credentials from AI infrastructure provider Fireworks AI. Up to 5,000 users who requested the affected versions may have received it. If your team uses Cursor, follow its remediation guidance, and install only editor extensions you need. [BleepingComputer]

UnrealIRCd chat-server flaw is being exploited
CISA has added a high-severity flaw in UnrealIRCd, an open-source Internet Relay Chat (IRC) server, to its Known Exploited Vulnerabilities catalog, which means attackers are using it. The Hacker News reports it is fixed in version 6.2.7.1. If you or your IT provider run it, update. Not sure? Ask your provider to check. [The Hacker News]

AI is making phishing more personal
Edwin’s CyberDWN feed points to CrowdStrike’s 2026 Global Threat Report, which shows attackers using AI to scale their campaigns. As advice, expect phishing that looks more personal and less sloppy. If a request involves money, passwords or bank-detail changes, confirm it through a second channel, such as a call to a number you already have. [CyberDWN feed]

Salesforce guide: AI guardrails need human oversight
A CyberDWN feed post covers a practical guide to AI guardrails and governance published by Salesforce. Its key idea is that guardrails are more than filters: they are a governance layer that needs human oversight. As advice, decide who reviews important AI output before it reaches a customer or your books, and write down who has the final say. [CyberDWN feed]

FAA ground stop: no sign of a cyberattack
SecurityWeek reports that officials say there is no evidence of a cyberattack behind the nationwide FAA ground stop on November 13. It was lifted at 8:15 pm ET, more than four hours later, and delays are expected for several days. The FBI is reportedly examining a software update. If you are travelling this week, allow extra time. [SecurityWeek]

Apple explains Secure Exclaves for camera and microphone data
Apple has detailed Secure Exclaves, which SecurityWeek says isolate ambient sensor data, such as camera and microphone data, from the main operating system on HomeView and new iPhones. In plain terms, that data is walled off from the rest of the system. Nothing urgent to do, but keep your devices updated so you get protections like this. [SecurityWeek]

Story of the week: Outlook zero-day patched
Microsoft has released emergency patches for the exploited Outlook spoofing flaw. This follows the November 8 daily recap, which last covered the story, and the November 14 recap reports the fixes are now out in the November 2026 Patch Tuesday updates. Exploited means attackers were already using the flaw, so install the patch promptly.
For a small business, the job is short. Make sure Outlook and Windows are set to take the November updates, and confirm your IT provider has applied them to every machine, including laptops that rarely visit the office. Keep treating unexpected emails with care, even when they appear to come from someone you know. [CyberDWN feed]

What to do this week
- Join video meetings only through the official app, and never run a fix someone sends you.
- Install the November Microsoft updates, and update UnrealIRCd to 6.2.7.1 if you run it.
- If your team uses Cursor, follow its remediation guidance.
- Confirm unexpected requests for money or logins through a second channel.
- Decide who reviews important AI output.
Sources
- https://www.bleepingcomputer.com/news/security/us-disrupts-north-korean-campaign-behind-400-million-crypto-theft/ (BleepingComputer, 2026-11-14)
- https://www.bleepingcomputer.com/news/security/cursor-confirms-security-breach-led-to-malicious-claude-code-extension/ (BleepingComputer, 2026-11-14)
- https://thehackernews.com/2026/11/attackers-exploit-unrealircd-flaw-that.html (The Hacker News, 2026-11-14)
- https://edwinvonlabja.com/2026/11/14/ai-driven-cyber-threats-countermoves-2/ (CyberDWN feed, 2026-11-14)
- https://edwinvonlabja.com/2026/11/14/ai-guardrails-governance-small-teams/ (CyberDWN feed, 2026-11-14)
- https://www.securityweek.com/no-signs-of-hacking-in-faa-ground-stop-as-fbi-reportedly-probes-software-update/ (SecurityWeek, 2026-11-14)
- https://www.securityweek.com/apple-reveals-how-secure-exclaves-shield-camera-and-microphone-data/ (SecurityWeek, 2026-11-14)
- https://edwinvonlabja.com/2026/11/14/daily-cyber-recap-20261114/ (CyberDWN feed, 2026-11-14)
Follow for the daily recap
Learn more at edwinvonlabja.com, and watch the recap on YouTube: https://www.youtube.com/watch?v=Tn41QXoOEWo
Produced with an AI-assisted pipeline; every claim is checked against the linked sources.
