Weekly AI & Cybersecurity Digest: Black October Hits AI Developer Platforms, OpenAI’s 948 Root Exploits, and CrowdStrike’s AI Kill Switch

1️⃣ Breaking News

Black October: A Dozen AI Developer Platforms Breached for Their Customers’ AI Keys

At least a dozen AI developer platforms disclosed breaches in a month, most recently Northflank, Trigger.dev, Chainstack and WorkOS. Attackers hunted stored environment variables holding AI provider keys, then resold or abused them. Wiz saw stolen enterprise keys tested within 30 minutes of a vendor breach. Entry points differed and the attacks do not appear linked. If a platform stores your AI keys, rotate them and cap spending now. (SecurityWeek, Wiz)

WordPress Core Flaw CVE-2026-119977 Added to CISA’s Exploited List

CISA added CVE-2026-119977, a time-of-check time-of-use race condition in WordPress Core, to its Known Exploited Vulnerabilities catalog on November 3, based on evidence of active exploitation. Versions before 7.1.4 are affected. WordPress shipped 7.1.4 as a security release on October 28 and urges immediate updates. Sites with automatic background updates should already be patched, but it is worth checking your version today. (CISA, WordPress)

Poisoned cPGuard Update Ran as Root on About 2,950 Hosting Servers

On November 2, an attacker pushed fake cPGuard updates (versions 5.88.12 to 5.88.19) through OPSShield’s own update channel. For over 70 minutes, about 2,950 customer servers installed them and ran a script as root that stole password hashes, root SSH keys and cloud storage keys. OPSShield says the attacker first signed up as a customer. Affected admins should update to 5.88.26 and rotate every exposed secret. (OPSShield, Help Net Security)

2️⃣ Research Highlights

OpenAI’s GPT-6 Cyber Produced 948 Linux Root Exploits in One Run

In a mid-October run, OpenAI’s pipeline reproduced 1,724 Linux kernel vulnerabilities and produced 948 distinct privilege escalation exploits, at $1,279 per exploit in API credits. Patches were developed for each and are being sent to maintainers. On November 5, OpenAI announced the model, GPT-6 Cyber, with access restricted to vetted defenders. Its warning: for a well-resourced attacker, a low-privileged account on an unpatched system is now as good as root. (OpenAI, BleepingComputer, iTnews)

UK NCSC: AI-Powered Ransomware Against Critical Infrastructure “Highly Likely” by 2028

The NCSC’s State of the Cyber Nation report, published November 3, says a ransomware attack on UK critical national infrastructure using AI-powered malware is “highly likely” by 2028. CEO Richard Horne argues AI is not inventing new attacks; it exposes long-tolerated weaknesses at unprecedented speed. His message to leaders: the era of “tech on the cheap” is over, and resilience must be funded now. (NCSC)

3️⃣ Featured Tools & Resources

CrowdStrike Launches an AI Kill Switch and Buys Fiddler AI and Dreadnode

At Fal.Con Europe, CrowdStrike launched an AI Kill Switch in Falcon Guardian that lets an authorized administrator stop a rogue agent, block it from restarting and cut its network access. For now, it blocks agents’ access to AI models at the gateway; endpoint, cloud and identity containment are planned. CrowdStrike also agreed to buy Fiddler AI and Dreadnode, for a combined $495 million according to the WSJ. (CrowdStrike, Yahoo Finance)

4️⃣ Bonus: Emerging Threats or Industry Events

European Supermarket Fridges Hacked Again, With Freezers Set to 50 °C

On November 2, supermarkets in at least six European countries, including Sweden, Belgium, Luxembourg, Ireland, Italy and Estonia, reported fridges and freezers warming after remote takeovers. Dozens of stores were hit and tons of food thrown away; one Swedish store found its units set to 50 °C. French investigators blame outdated Danfoss and Copeland remote controllers; experts cite end-of-life firmware and default passwords. (AP News, CBS News)

Zyxel GS1900 Zero-Day Bypassed a June Patch and Hit Nearly 1,000 Switches

Attackers compromised 996 Zyxel GS1900 switches in 48 countries using CVE-2026-107441, an unauthenticated out-of-bounds write that walked past June’s fix for CVE-2026-7273. They stole configuration files and credentials. Zyxel shipped fixed firmware for ten models on October 22 and published its advisory on November 3. Patch now, keep switch management interfaces off the internet, and rotate any credentials stored on affected devices. (BleepingComputer, Secure.com)


Stay informed and vigilant as the fields of AI and cybersecurity continue to evolve rapidly.


Discover more from Science & Tech

Subscribe to get the latest posts sent to your email.

Rating: 1 out of 5.

Leave a Reply

Get updates

Whether you’re a seasoned professional or just someone passionate about the intersection of science and technology, there’s something here for you, all here in our weekly newsletter.

Adversarial Attacks Agentic AI Security Tools AI AI Regulation and Ethics AI Security AI Security 2025 AI Vulnerabilities Automation Autonomous AI Agent Risks Autonomous AI agents Awareness Breaches CISO Cloud Compliance Credentials Critical Infrastructure Culture Cybercrime Cybersecurity Cybersecurity News Emerging Cyber Threats Ethic Hacking Infosec Large Language Model Risks Leadership Least Privilege Prompt Injection Attacks Regulations Resilience Risk Management SOAR Social Engineering SupplyChain Threat Detection Threat Intelligence Threats Threats Management Training Trends Vulnerability XDR Zero-Day Exploits Zero-Trust

Last posts (articles)

Disclaimer: Web links are not guaranteed to be up-to-date.

Archives (Articles)

Archives (Podcasts)

You can also find our podcast on these streaming services (and many more):