Week in review: Secret stealer hits GitHub repos, Claude Code turned into remote-controlled implant

Here’s an overview of some of last week’s most interesting news, articles, interviews and videos:

What a security leader should tell the CEO about leaked credentials
In this Help Net Security interview, Eric Fourrier, CEO of GitGuardian, talks about leaked credentials and personal liability for security leaders and boards. He explains how fear of blame pushes mistakes out of sight, and what a security leader should be ready to tell a CEO about exposed keys.

Turla’s Cereal malware turns victims’ Exchange servers into spy hubs
Russia-aligned group Turla planted a modular .NET toolset on Microsoft Exchange servers at two government organizations, one in Kazakhstan and one in Romania, and used it to copy selected email straight out of the mail flow. ESET researchers found the malware and named it Cereal.

AI reverse engineering let researchers add a new feature to Slack
Intezer researchers used AI agents to rebuild source code for Slack, Apple’s Finder and a TinyTurla-NG sample, from a backdoor family Cisco Talos attributes to the Turla group. They worked from the shipped programs and vendor documentation. The rebuilt Slack ran well enough that the team added a new feature to it: AI text autocomplete, powered by a small local model.

Nearly half of big firms shelve or close vulnerability findings they haven’t confirmed
Nearly half of security teams at organizations with 5,000 or more employees close or deprioritize vulnerability findings without fully verifying them, ProjectDiscovery found, and the reason they give is time.

EU cyber scores look healthy, but 45% of surveyed IT providers take months to patch
Forty-five percent of surveyed IT service management providers in the EU took more than three months to apply critical patches to critical systems, according to ENISA’s 2026 assessment of cybersecurity in the Union. Attackers can turn a newly disclosed flaw into a working attack within days, and sometimes within hours or minutes.

The Cyber Resilience Act clock doesn’t stop for weekends
In this Help Net Security interview, Aurimas Bakas, CEO of Copla, talks about what the EU Cyber Resilience Act‘s reporting duties mean for manufacturers. He explains why companies need a register of every product version still in use, why boards focus on CE marking while reporting obligations started in September 2026, and how to define the moment a company becomes aware of an exploited flaw.

CISOs say OT will survive a cyberattack, but their engineers aren’t so sure
Ask industrial security chiefs whether their plants would keep running through a full IT compromise, and every one who answered the SANS 2026 ICS/OT survey says yes. Thirty-eight percent expect to run indefinitely on islanded ICS/OT systems, and 62% expect to get by temporarily on manual controls. None foresee degradation or a halt.

Dagster: Open-source cloud-native data pipeline orchestrator
Dagster is an open source orchestrator for data pipelines. It decides what runs, in what order, and when. It also tracks how the tables, datasets, machine learning models, and reports it manages connect to the data they came from.

GitHub Agentic Autofix and the question of who owns a bad AI patch
In this Help Net Security interview, Marcelo Oliveira, VP of Product Management at GitHub, talks about who is accountable when an AI-generated patch fixes one flaw and introduces another. He explains how Agentic Autofix re-runs detection engines and project tests, and what a CISO should change before pointing an agent at thousands of old alerts.

Check your GitHub repos: A secret stealer hit tens of thousands of them
A threat actor using stolen GitHub credentials has pushed a credential-stealing payload into tens of thousands of repositories since November 10. In more than 1,800 of them, the malware ran and shipped GitHub Actions secrets and cloud credentials to the attacker. Aikido spotted the campaign and named it wfx-loot.

Stolen Claude API keys can turn Claude Code into a remote-controlled implant
Darktrace researchers turned Claude Code into a remote-controlled implant in a lab test, steering the coding agent through the memory stores feature of Anthropic’s Claude platform and pulling a 1 GB file out through the Claude Files API.

Operation Endgame malware takedown turns up over 6 million stolen logins
Police in nine countries, working with Europol, took down servers behind the Gootloader, CastleRAT and StealC malware operations this week and recovered more than 6 million credentials stolen from infected computers.

Attackers breached 10 organizations with an open-source AI pentest framework
Chinese-speaking attackers used Cairn, an open-source AI penetration testing framework, to run 25 operations against at least 19 organizations in August and September 2026. Palo Alto Networks’ Unit 42 found that the attackers gained unauthorized access or pulled data at 10 of them, often after typing little more than a domain name and a one-line goal.

Fake Zillow pages use ClickFix NetSupport RAT to target apartment hunters
EfficientIP Research Labs traced a ClickFix campaign that used fake Zillow apartment-review pages to install NetSupport RAT on visitors’ Windows computers. Each page showed what looked like a routine verification check. Behind it, the page quietly copied a PowerShell command to the clipboard and walked the visitor through pasting and running it.

Your pinned MCP server may not be as pinned as you think
An MCP server configured to run one fixed release on an administrator’s workstation ran with 15 different sets of dependencies over five weeks, with its configuration untouched. James Lagermann, Director of Technical Alliances at Corelight, found the drift in his company’s lab.

Suspected North Korean IT workers reached onboarding at companies with open data
Material Security followed suspected North Korean fake IT workers through the hiring pipelines of organizations it monitors and counted 27 personas that reached onboarding over one year. At one company where a persona got that far in the past month, about 277,000 files containing Social Security numbers were shared company-wide or by public link.

How attackers turned a Teams meeting booking into a device code phishing trap
Attackers used a fake Microsoft Teams meeting scheduler to trick an employee at a construction company into approving a device code sign-in. They got a valid session token before Huntress’s security operations center identified the incident.

Stolen webcam clips of 900+ victims feed North Korea’s $400 million fake Zoom scheme
North Korean state hackers have stolen more than $400 million in cryptocurrency over the past 18 months by pulling targets into fake Zoom and Teams calls filled with webcam footage of earlier victims. The FBI calls the scheme FaceStealer.

354 dark web vendors and buyers arrested in Operation TorNado
Europol and police in 14 countries arrested 354 dark web vendors and buyers in Operation TorNado, which ran from May 2025 to October 2026. Investigators went after people who had traded on five marketplaces that were already offline: Archetyp, Bohemia, Nemesis, Kingdom and Tor2Door.

Exploited SonicWall SSL-VPN flaw lets attackers bypass SAML-enforced MFA (CVE-2026-19868)
SonicWall has fixed an exploited SonicOS SSL-VPN vulnerability (CVE-2026-19868) that may allow attackers with valid credentials to bypass SAML-enforced multi-factor authentication. With no workarounds listed, the company has urged admins to patch quickly.

ARTEX goes from pentesting tool to attacker staple in three months
A security researcher’s AI agent breached an internal sales application used by WorkOS’s go-to-market teams, the developer platform provider confirmed on Friday. The researcher confirmed that they deleted and did not further share the customer data the agent obtained from WorkOS’s systems, and the company acknowledged the security failings that led to the breach and its delayed detection of the unauthorized API requests and data retrieval.

Photos: IRISSCON 2026 cybercrime conference
IRISSCON 2026, the annual cybercrime conference organized by IRISSCERT, took place at the Aviva Stadium in Dublin on November 11. The all-day event gives attendees a look at the cyber threats businesses are facing right now, along with practical advice on how organizations can respond to them.

A CISO’s guide to trimming waste from a flat security budget
In this Help Net Security video, Ken Leeser, CISO at Y Meadows, explains how security leaders can get more protection out of a flat budget. He argues a frozen budget gives teams a reason to cut waste and simplify their security stack.

Who collects the keys when an AI agent retires?
Deleting an agent from a dashboard can leave active credentials, unfinished tasks, and workflows that still depend on it. It can also erase the context needed to explain decisions long after the agent is gone. Girish Redekar, CEO at Sprinto, explains why AI agents need formal offboarding, and how organizations can revoke access, hand over work, preserve decision records, and assign clear ownership as their agent populations grow.

$133 million stolen from French bank via exploited service provider
Attackers siphoned an estimated $133 million from Crédit Mutuel Arkéa by using the bank’s own account to wire money to other banks, from where it was possibly converted into cryptocurrency.

Hackers abuse refrigeration controllers to send spam and route proxy traffic
Internet-exposed refrigeration controllers used in supermarkets and cold storage facilities are being hijacked and turned into infrastructure for a suspected spam botnet and proxy services, SecurityScorecard has found.

An average of 8 minutes is all hackers need to log in with stolen credentials
Attackers used stolen credentials to log in an average of eight minutes after infection across 27 confirmed infostealer cases, according to a report from Elastic Security Threat Command. Researchers also documented misuse of AI applications to execute commands and digitally signed remote monitoring tools to maintain access.

Companies seeking an MSP doubt their ability to spot cyberattacks
Businesses looking to outsource cybersecurity want providers that can help them recover from attacks, manage AI risks, and explain what security problems mean for their operations, according to Barracuda What MSP Customers Need Next report.

Wirken: Open-source AI agent gateway
Wirken is an open-source gateway that sits between messaging apps and AI agents. It assumes any agent can be turned against its owner and limits what a compromised one can reach.

Product showcase: What the Malwarebytes Claude connector reveals about unexpected texts
The Malwarebytes connector for Claude checks suspicious links, phone numbers, and email addresses without leaving the chat. You can paste a link or upload a screenshot of a message, then ask Claude to check the details against Malwarebytes’ threat database.

PyPI makes release attestations easier to find with new security tab
The Python Package Index (PyPI) has added a security tab to its project pages, giving developers a place to check where individual release files came from. The Security tab shows whether release files have digital attestations, which provide verifiable information about how a file was published or built. A release has full provenance when every file has at least one attestation.

Over 575 malicious OpenClaw skills delivered malware to Windows and macOS
Cybercriminals are using AI agents to coordinate attacks, stolen credentials to access linked services, and tools that interfere with intrusion detection and system recovery, according to Acronis TRU Cyber Pulse 2026.

Security teams want to spot unusual AI agent behavior while they’re still defining normal
AI agents topped the list of planned investments in behavioral monitoring in Abnormal AI’s 2026 State of Behavioral Security Report. Asked where their organizations would invest most over the next two years, 43% of respondents chose AI agents, more than three times the share who chose API integrations, the next most common answer.

Windows settings backup is on by default starting with Windows 11 26H2
Starting with Windows 11, version 26H2, Windows settings backup will switch on by default on eligible devices. The company renamed the feature, which launched in August 2025 as Windows Backup for Organizations, and moved the management of roaming settings into the same policy tools.

Claude Mythos 5.2 arrives with stronger bug hunting and less sandbox-breaking behavior
Anthropic released Claude Fable 5.2 and Claude Mythos 5.2. The two versions share an underlying model, with stricter safeguards for cybersecurity and biology tasks in the broadly available Fable.

CISA tweaks cyber incident form to spot attacks that might be linked
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has updated its online incident reporting form to help analysts spot when attacks on different organizations may be part of the same campaign. The revised questionnaire asks more specific questions and removes duplicate ones.

X.Org patches eight server vulnerabilities a month after fixing 12 others
X.Org has fixed eight vulnerabilities in the X.Org X server and Xwayland, including one in the Xorg server that lets a malicious monitor corrupt its memory. The fixes ship in xorg-server 21.1.26 and xwayland 24.1.15, and each of the eight affects versions before those two releases.

Firefox free VPN drops its 50GB launch cap until January 31
Mozilla has removed the data limit on the free VPN built into Firefox. Firefox users in supported regions get unlimited browsing on desktop and Android once they sign in with a Mozilla account.

Google Fi super network leans on encryption to keep hotspot traffic private
Google Fi now moves customers’ phones onto Wi-Fi hotspots and between two 5G networks without the user going into settings or logging in. Google calls the setup a super network, and it powers all Fi plans on eligible Pixel phones.

An Instagram ad was all it took: Leaked exploit kits are hitting employees’ iPhones
Criminals are using leaked exploit kits to break into vulnerable iPhones, according to iVerify’s 2026 Advanced Mobile Threat Report. Attacks delivered through ads and compromised websites have reached employees who use phones for work.

Ubuntu certificate revocation checks arrive in curl, OpenSSL is next
Ubuntu 26.10 ships curl with certificate revocation checking built in. The check comes from upki, a project from the rustls developers. Ubuntu’s developers ran the same curl command on both releases: Ubuntu 26.04 accepted a revoked certificate and connected, while 26.10 refused it.

Microsoft researchers demonstrate hidden backdoor technique for AI agents
Microsoft researchers have demonstrated how small changes to an AI model can implant a hidden backdoor, causing an agent to execute an attacker’s command when a specific trigger appears. The technique, called Lazarus, requires no model retraining and allows the agent to behave normally during standard safety evaluations.

Git fixes six vulnerabilities, including code execution during recursive clones
The Git project has released security updates addressing six vulnerabilities that could allow attackers to execute code, write to unintended files or cause memory corruption when Git processes malicious repositories or other specially prepared input.

Bots pose as AI agents, AI agents pose as humans, and defenses believe both
Over one week in October, an online marketplace took 414,210 visits from traffic wearing the name of a widely used consumer AI agent. Every visit went for the sign-in page, and every one returned errors. Behind the name was a credential stuffing operation, with attackers running stolen username and password pairs against the site in bulk.

.NET 11 security changes target the software supply chain
Microsoft has released .NET 11, a Standard Term Support (STS) release that will receive two years of support. .NET 10 is still the Long Term Support (LTS) option. The release covers C# 15, ASP.NET Core, .NET MAUI, and Windows development. Several of the changes affect software supply chain integrity, authentication handling, and platform lifecycle.

AWS Emulator: Open-source tool for testing AWS apps locally, built by autonomous AI agents
Amazon Web Services has released AWS Emulator, a free tool that simulates AWS services on a developer’s laptop, a development server or in a continuous integration (CI) pipeline. The company says autonomous AI agents wrote the tool entirely, with no code written or reviewed by its engineers.

Microsoft Entra adds a seven-day safeguard against irreversible identity deletion
Microsoft Entra’s Hard Deletion Protection, now in preview, gives administrators a seven-day window to recover selected identities before they can be permanently erased. The feature is designed to help organizations respond to accidental or malicious deletion of critical directory objects.

Self-hosted Claude Managed Agents keeps agent credentials in your environment
Anthropic has released a self-hosted version of Claude Managed Agents. Companies can run the agent harness, sandboxes, credentials, and session history in their cloud or on-premise. The option opened as a private beta on November 12, and customers request access through their Anthropic account team.

PostgreSQL security update fixes 52 vulnerabilities, ends support for PostgreSQL 14
The PostgreSQL Global Development Group released PostgreSQL 18.7, 17.12, 16.16, 15.20 and 14.25 on November 12, fixing 52 security vulnerabilities and more than 140 other bugs.

Google smart glasses privacy features cover wearers and the people around them
Google has built its Gemini-powered intelligent eyewear to stop taking photos and video when someone covers or tampers with the outward-facing recording light. Shahram Izadi, Google’s VP and GM of XR, laid out that safeguard along with face redaction, limits on data handling, and app restrictions.

Digital sovereignty leaves enterprises chasing their own audit trails
94% of surveyed enterprises experienced at least one digital sovereignty challenge in the past three years, according to IBM’s Digital Sovereignty Index. Of those affected, 91% were required to demonstrate that they retained control during their most challenging event, and 81% had to provide formal evidence to regulators, auditors, customers or internal stakeholders.

OpenAI’s Daybreak brings AI agents to security testing and vulnerability fixes
OpenAI is bringing AI agents into more of the software security process with Daybreak, pairing web application penetration testing with tools that investigate vulnerabilities and develop fixes. The two offerings, Red Team and Defense Factory, connect the work of finding security weaknesses with patching and retesting them.

Apple details Secure Exclaves, isolated runtimes for camera and microphone data
Apple published details of Secure Exclaves, a security architecture designed to protect camera and microphone data during background processing. Apple says data processed and analyzed within Secure Exclaves cannot be accessed by the main operating system, apps, or the company itself.

Cybersecurity jobs available right now: November 10, 2026
We’ve scoured the market to bring you a selection of roles that span various skill levels within the cybersecurity field. Check out this weekly selection of cybersecurity jobs available right now.