Watch the 2-minute recap
Google says attackers are now using autonomous AI agents to find and exploit software flaws at machine speed. In the last day or so, reports also covered a suspected supply chain attack on news sites, a 22 million-person address exposure in France and a new Barracuda breach detail. Here is what it means for small businesses.
Today at a glance
- Attackers now use autonomous AI agents. [The Hacker News]
- Project Zero shortens bug disclosure deadlines. [SecurityWeek]
- Suspected supply chain attack hits major sites. [BleepingComputer]
- Maileva exposes 22 million people's addresses. [Infosecurity Magazine]
- Telegram widgets hit by code execution demos. [SecurityWeek]
- BO Team uses Log4Shell on a Russian firm. [The Hacker News]
- Coles investigates a cyber incident. [BleepingComputer]
- Barracuda's breach used a stolen key and Artex. [BleepingComputer]
Google says attackers now use AI agents
Google says threat actors are now using autonomous, multi-agent offensive frameworks, named ARTEX and CAIRN among others, to carry out multi-stage intrusions. Reports say these tools automate finding and exploiting vulnerabilities, which speeds attacks up to machine pace. Dark Reading reports China-nexus actors used Cairn against 10 organizations. If you run a business, treat patching as routine, not optional: apply updates soon after release and retire software you no longer need. [The Hacker News]

Project Zero shortens its bug deadlines
Project Zero, Google’s long-running bug-hunting team, will move to a 60-day disclosure window in February 2027 and to a 30-day window in January 2028. Reports say the change responds to AI-accelerated vulnerability discovery and exploitation. Vendors get less time to fix a flaw before details go public. For you, that suggests turning on automatic updates wherever you can, so fixes land as soon as they ship. [SecurityWeek]

Optimizely supply chain attack hits big sites
BleepingComputer reports that a suspected supply chain attack targeting Optimizely led to malicious JavaScript being loaded on numerous popular websites, including CNN, Fox News, BBC and Salesforce. The scripts were designed to steal cryptocurrency and payment information. A supply chain attack means the trouble came through a trusted outside provider rather than the sites themselves. If your website loads third-party scripts, list them, remove any you do not need and ask each vendor how they are secured. [BleepingComputer]

Maileva breach exposes 22 million people
Infosecurity Magazine reports that a breach at Maileva, a subsidiary of the French postal group La Poste, exposed the names and postal addresses of 22 million people. Names and addresses can make scam letters and phishing messages look convincing. If you use a postal or mailing service, verify any unexpected request before you act on it. [Infosecurity Magazine]

Telegram widgets hit by remote code execution demos
SecurityWeek reports that researchers demonstrated remote code execution on Telegram’s new widgets just days after launch. Remote code execution means running an attacker’s code on someone else’s device. Kaspersky suspects the widgets are being exploited to deliver WebRAT, while Pavel Durov disputes the claims. The picture is still unsettled, so keep Telegram updated, be careful with unexpected widgets or links, and follow official guidance as it appears. [SecurityWeek]

BO Team uses Log4Shell against a Russian firm
The Hacker News reports that a group called BO Team exploited Log4Shell in VMware vCenter and compromised a Russian firm’s code-signing server. The report describes a months-long compromise of the developer’s infrastructure, documented by Kaspersky’s ICS CERT team. A code-signing server is what vouches that software is genuine, so losing control of it is serious. As advice, patch older, well-known flaws and keep management tools off the open internet. [The Hacker News]

Coles investigates a cyber security incident
BleepingComputer reports that Coles Group, Australia’s second-largest retailer, is investigating what it described as a cyber security incident that has not impacted the company’s stores. Few details are public so far. Customers and suppliers should watch for official updates and treat unexpected emails that mention the incident with care until more is confirmed. [BleepingComputer]

Story of the week: Barracuda's cloud breach and Artex
BleepingComputer reports that Barracuda, the network and email security company, says its cloud breach used a stolen cloud access key and the ARTEX AI penetration testing framework. This follows our November 16 recap, which covered the FBI warning about the ARTEX AI hacking tool.
It ties neatly to today’s top story: attackers pairing stolen credentials with automated tooling. A cloud access key works like a password for your cloud account, so treat one as a valuable secret. As advice, rotate keys on a schedule, remove keys nobody uses, limit what each key can do and turn on alerts for unusual activity. Even a security company can be caught, so build habits that assume a key could leak, and review who can create keys. [BleepingComputer]

What to do this week
- Turn on automatic updates and patch anything internet-facing as soon as fixes appear.
- List every third-party script on your website and remove the ones you no longer need.
- Rotate cloud access keys and delete any that are unused.
- Treat unexpected mail, texts and links with care, even when they look polished.
- Keep an offline backup and test that you can restore from it.
Sources
- The Hacker News: https://thehackernews.com/2026/11/google-finds-hackers-using-ai-agents-to.html (2026-11-16)
- BleepingComputer: https://www.bleepingcomputer.com/news/security/google-warns-that-fully-automated-ai-attacks-are-coming-soon/ (2026-11-16)
- Dark Reading: https://www.darkreading.com/cyberattacks-data-breaches/china-nexus-actors-breach-10-organizations-ai-driven-attacks (2026-11-16)
- SecurityWeek: https://www.securityweek.com/google-project-zero-sets-course-for-30-day-disclosure-deadline/ (2026-11-16)
- BleepingComputer: https://www.bleepingcomputer.com/news/security/google-project-zero-cuts-bug-disclosure-deadline-to-30-days-amid-ai-threats/ (2026-11-16)
- Dark Reading: https://www.darkreading.com/vulnerabilities-threats/google-project-zero-bug-disclosure-deadlines (2026-11-16)
- BleepingComputer: https://www.bleepingcomputer.com/news/security/optimizely-supply-chain-attack-delivers-malware-to-cnn-fox-news-bbc-salesforce-sites/ (2026-11-17)
- Infosecurity Magazine: https://www.infosecurity-magazine.com/news/maileva-docaposte-22-million-breach/ (2026-11-16)
- SecurityWeek: https://www.securityweek.com/telegram-widgets-hit-by-rce-demos-malware-warning-days-after-launch/ (2026-11-16)
- The Hacker News: https://thehackernews.com/2026/11/bo-team-exploits-log4shell-in-vmware.html (2026-11-16)
- BleepingComputer: https://www.bleepingcomputer.com/news/security/australian-retail-giant-coles-investigates-cybersecurity-incident/ (2026-11-16)
- BleepingComputer: https://www.bleepingcomputer.com/news/security/barracuda-links-cloud-breach-to-artex-ai-powered-cyberattack/ (2026-11-16)
Follow for the daily recap
Follow if you like this content and want to stay up to date. Visit edwinvonlabja.com or watch on YouTube: https://www.youtube.com/watch?v=f_OczDJVh6g
Produced with an AI-assisted pipeline; every claim is checked against the linked sources.
