News
Every CyberSIXT summary, newest first, with a link to its story and to the original report.
Most articles on Wed 21 Oct (96); 10–13 on weekend days.
- On CISA KEV CVEs
- Other coverage
Show as table
| Day | Articles published |
|---|---|
| Mon 12 Oct | 56 |
| Tue 13 Oct | 85 |
| Wed 14 Oct | 86 |
| Thu 15 Oct | 78 |
| Fri 16 Oct | 61 |
| Sat 17 Oct | 10 |
| Sun 18 Oct | 10 |
| Mon 19 Oct | 60 |
| Tue 20 Oct | 86 |
| Wed 21 Oct | 96 |
| Thu 22 Oct | 84 |
| Fri 23 Oct | 68 |
| Sat 24 Oct | 13 |
| Sun 25 Oct | 13 |
| Today, Mon 26 Oct | 59 so far |
Sun 4 Oct· 19 articles
-
Google Restricts Free Gemini Users to Its Basic Flash-Lite Model
Google is restricting access to Gemini AI models for free personal accounts, with the change taking effect on 9 October 2026. Under the new policy, users on free personal accounts will be limited to the Gemini Flash-Lite model, while higher-end models such…
SecurityOnline -
Meta Opens Muse AI to Makers With Hardware Development Push
Meta has expanded its Muse AI strategy by inviting the maker community to design hardware for Muse, launching Muse Gadgets alongside its Muse Charm demo at Connect. The initiative aims to move Muse beyond phones and wearables, positioning it as a platform…
SecurityOnline
Sat 3 Oct· 20 articles
-
Italy Fines IQVIA €7 Million Over Identifiable Patient Health Data
The Italian Data Protection Authority has fined IQVIA Solutions Italy Srl €7 million following an investigation into a data protection breach. The authority found that IQVIA had created a database containing health information on around one million patients…
DataBreaches -
ShinyHunters Hacker Detained in Jordan and Reportedly Aids FBI Investigation
Saif al-Din Khader, known online as “Rey” and “Hikki-Chan”, a key member of the ShinyHunters hacking group, has been detained by Jordanian authorities, according to three people familiar with the matter who spoke to Reuters. The sources said Khader was…
DataBreachesShinyHunters -
Fake Zoom Installer Steals macOS Passwords to Deploy CloudSyncD
Jamf Threat Labs uncovered CloudSyncD, a macOS backdoor hidden inside a fake Zoom installer. The malicious package presents as a user-facing macOS installer, with the disk image mounted as a volume named Zoom and an Applications shortcut. It guides the user…
Security Affairs -
YARA-X 1.21.0 Streamlines Automated Scans With Stdin Support
The Internet Storm Center reports that YARA-X has released version 1.21.0, bringing five improvements and four bug fixes. The update appears to focus on enhancing usability and reliability for users who employ YARA-X for scanning tasks, with specifics likely…
SANS ISC -
MI5 Warns China Funded UK Research Through Academic Front
The UK’s MI5 has issued a Security Service Espionage Alert alleging that China’s MSS has funded research through the China General Technology Research Institute (CGTRI), also known as the China Academy of General Technology (CAGT). The agency says CGTRI is a…
The Hacker NewsMSS -
China-linked Warlock Exploits SharePoint to Deploy Ransomware Against Critical Sectors
Warlock, a China-linked threat actor also known as Gold Salem, Longlegs and Storm-2603, continues to weaponise flaws in Microsoft SharePoint to gain initial access, disable security tooling and deploy ransomware. The activity, tracked by Symantec and Carbon…
-
DHS Plans $626m Contract to Centralise Cybersecurity and Cloud Services
The Department of Homeland Security is preparing to award a major new contract aimed at centralising and tightening the delivery of IT services across the department. The NCCS 2.0 vehicle—standing for Network, Cloud and Cybersecurity Services 2.0—will cover…
DataBreaches -
California AG Subpoenas OpenAI Over Frontier Model Security Risks
OpenAI is facing a formal investigative demand from California’s Attorney General, who served OpenAI with an investigative subpoena aimed at obtaining more detail on the security of its frontier models and the risks they pose. The move comes as regulator…
DataBreaches -
Federal Reserve Employee Removed Sensitive Files Before Retirement
A Federal Reserve Board employee in the Division of International Finance repeatedly removed sensitive files before planned retirement, according to the Federal Reserve’s inspector general (OIG). The issue came to light during an audit of the Fed’s…
DataBreaches -
Over 543,000 Credentials Lingered in Public GitHub Repositories
More than 543,000 valid credentials were exposed in public GitHub repositories, according to research cited by Dissent and Bill Toulas. The study, conducted by Truffle Security, analysed data pulled from scanning 224 million repositories and more than 58…
DataBreaches -
US Senate Advances Bipartisan Bill to Strengthen Healthcare Cybersecurity
The U.S. Senate has advanced bipartisan legislation aimed at strengthening cybersecurity in the healthcare sector. By unanimous consent, senators approved the Health Care Cybersecurity and Resilience Act, a move described by officials as a means to help…
DataBreaches -
Epic Pauses Development After MyChart Flaw Risks Unlogged Patient Access
Epic, the software giant behind the MyChart patient portal, has paused most of its product development to focus on addressing security flaws that could put patients’ data at risk. The move, described as likely to last about six weeks, follows findings from…
DataBreaches -
MetaMask Isolates Staking Validators Amid Ongoing Infrastructure Incident
Metamask has disclosed an ongoing infrastructure security incident that is affecting some of its underlying systems. The wallet provider says it is working to address the issue with internal teams, external partners and security advisers, and emphasises that…
DataBreaches -
Doxx.net Raises $38 Million to Build a Private Internet for AI Agents
Doxx[.]net has announced a $38 million Series A round led by Andreessen Horowitz, with participation from Animo Ventures and Focal[.]vc, aimed at advancing its Agentic Defined Networking (ADN) platform. The Miami-based startup, founded in 2025 by Barrett…
SecurityWeek -
Fortra Patches Critical BoKS Flaw Enabling AD Password Cracking
Fortra has issued patches addressing eight vulnerabilities in its Core Privileged Access Manager (BoKS), including three rated as critical. The most severe, CVE-2026-79901 (CVSS 9.9), affects BoKS Manager deployments that use BoKS keytabs for Active…
-
2026 Security Outlook: Organisations Struggle to Control Expanding Attack Surfaces
The Hacker News reports on a 2026 cybersecurity outlook that frames security as increasingly cross-domain, with cloud, AI, and distributed systems expanding the number of identities, devices, data, and internet-facing infrastructure that organisations must…
The Hacker News -
GitLab AI Gateway Flaw Lets Users Run Commands on Hosts
GitLab has patched a critical vulnerability in its AI Gateway, CVE-2026-90970, rated 9.9 on the CVSS scale. The flaw could allow an authenticated user with access to the Duo Agent Platform to escape the prompt sandbox and execute arbitrary commands on the AI…
-
Rust Backdoor Uses Microsoft 365 to Spy on Asian Governments
Cisco Talos has been tracking a cluster of espionage activity since September 2025, named UAT-11587, which by July 2026 had affected at least 16 government and policy organisations across eight Asian countries. The operation centres on a Rust-written Windows…
Security AffairsUAT-11587 -
Apache OpenOffice Flaw Enables Code Execution Through Crafted Documents
Eight CVEs were disclosed by the Apache Software Foundation on 2 October 2026, spanning Apache OpenOffice, the Apache Directory LDAP API, and Apache Traffic Server. The most urgent issue is CVE-2026-59265 in Apache OpenOffice, which can trigger arbitrary…
-
Microsoft Reissues Exchange Updates With New CVE-2026-96940 Flaw
Microsoft has reissued the September 2026 security updates for Exchange Server as a V2 release, adding one new vulnerability: CVE-2026-96940. The Exchange Team says the update was published ahead of schedule and urges admins to apply it promptly. Microsoft…
Fri 2 Oct· 66 articles
-
Apple Tightens macOS Privacy Rules After Muse AI Message Access
Apple says it will tighten macOS privacy settings to curb misuse of Full Disk Access (FDA) permissions by third‑party AI agents, in a move following the Muse incident and a broader debate about how AI assistants handle private data. The change comes after…
Ars Technica -
Zammad2 articles this day · CISA first
Zammad Flaws Let Attackers Reach Root in Seconds, CISA Warns · CISA Flags Actively Exploited Zammad Flaw Enabling Remote Code Execution
-
Zammad Flaws Let Attackers Reach Root in Seconds, CISA Warns
U.S. CISA has added two Zammad GmbH Zammad flaws to its Known Exploited Vulnerabilities catalog. The vulnerabilities are CVE-2026-102489 (session fixation) and CVE-2026-102490 (improper privilege management), each with a CVSS of 9.4. CVE-2026-102489 affects…
-
CISA Flags Actively Exploited Zammad Flaw Enabling Remote Code Execution
CISA has added CVE-2026-102489 to its Known Exploited Vulnerabilities (KEV) catalogue. The vulnerability affects Zammad GmbH’s Zammad customer support platform and is known as the Zammad Session Fixation Vulnerability. It can lead to remote code execution as…
-
-
AWS Patches Critical Loom Flaw Allowing Full Admin Takeover
AWS has published two security bulletins on 2 October 2026 detailing four CVEs that affect Loom for AWS and SageMaker Unified Studio. Three flaws hit Loom for AWS, including a critical admin takeover vulnerability, while the fourth could allow a SageMaker…
-
Digi Routers Exposed to Root Command Injection via Web Interface
Digi International has disclosed a critical vulnerability in its Digi DAL OS (DAL OS) that affects the IX, EX and TX family of routers, Connect IT and Connect EZ devices, AnywhereUSB hubs, and XBee gateways. The flaw, tracked as CVE-2026-75937, enables an…
-
Chrome Update Fixes Critical WebGL Flaw Enabling Sandbox Escape
Google has released Chrome 154.0.8037.97 (and 154.0.8037.98 on the Stable channel) to address a raft of 11 security bugs, led by a critical WebGL flaw that could allow remote code execution outside the browser sandbox. The bug, CVE-2026-103628, is an…
-
GitLab AI Gateway2 articles this day · The Hacker News first
GitLab Patches Critical AI Gateway Flaw Enabling Command Execution · GitLab Fixes Critical AI Gateway Flaw Enabling Command Execution
-
GitLab Patches Critical AI Gateway Flaw Enabling Command Execution
GitLab has patched a critical flaw in its AI Gateway (CVE-2026-90970) that could allow an authenticated Duo Agent Platform user to execute arbitrary commands on self-hosted gateways. GitLab AI Gateway versions 18.1.6 up to, but not including, 19.2.4; 19.3…
-
GitLab Fixes Critical AI Gateway Flaw Enabling Command Execution
GitLab has patched a critical vulnerability in its AI Gateway that could allow a logged-in user with Duo Agent Platform access to execute arbitrary commands on the gateway under certain conditions. The flaw stems from a weakness in the prompt template…
-
-
Citrix NetScaler ADC2 articles this day · Dark Reading first
Citrix NetScaler Reboots Persist as SAML Flaw Fix Remains Pending · Citrix NetScaler Zero Days Exploited as Vendors Take Different Risks
-
Citrix NetScaler Reboots Persist as SAML Flaw Fix Remains Pending
Citrix NetScaler ADC appliances are reportedly rebooting repeatedly after patches were applied, even on builds 14.1-73.37. Citrix has acknowledged a new issue linked to NetScaler SAML authentication and is preparing a security bulletin and a fix, though no…
-
Citrix NetScaler Zero Days Exploited as Vendors Take Different Risks
Threat researchers tracked a sequence of zero-day events around Citrix NetScaler over the weekend of 24–26 September. GreyNoise Intelligence observed a US-based IP address scanning NetScaler installations and attempting remote code execution, with social…
-
-
RemoteThreat Raises $7 Million to Mimic Nation-State Hackers
RemoteThreat is an offensive cyber operations startup that aims to push red teaming beyond conventional methods by simulating attackers’ increasingly advanced capabilities. Founded in 2025 by Chris Thompson and Shawn Jones, the venture emerged from stealth…
Dark Reading -
China-Linked Antino Backdoor Targets Asian Government Bodies via Microsoft 365
Cisco Talos has disclosed a China-nexus espionage operation targeting government and policy bodies across eight Asian nations, including Taiwan, India, the Philippines, Cambodia, Pakistan, Thailand, and Myanmar. The group’s backdoor, codenamed Antino and…
The Hacker NewsUAT-11587 -
CISA Warns of Actively Exploited Zammad Flaw Enabling Root Access
CISA has added CVE-2026-102490 to its Known Exploited Vulnerabilities (KEV) catalogue. The vulnerability affects Zammad GmbH’s Zammad product and is identified as the “Zammad GmbH Zammad Improper Privilege Management Vulnerability”. It allows the local…
-
Dell Fixes Critical CSM Flaws Letting Attackers Seize Kubernetes Clusters
Dell has released security updates for Dell Container Storage Modules (CSM) to fix a set of critical flaws that could let unauthenticated attackers seize control of storage infrastructure and Kubernetes clusters. The vendor’s advisory lists six CVEs with…
-
Critical SConnect Flaw Could Expose SWIFT Systems to Remote Attacks
Researchers have disclosed a critical vulnerability in SConnect, a hardware-authentication middleware component used to access highly sensitive global government and financial systems, including the SWIFT network. The flaw resides in the browser extension…
-
Cloudflare’s Streamline Brings Real-Time Video Editing to Workers
Cloudflare’s Streamline introduces a developer playground for building bespoke video pipelines on Cloudflare’s platform. It demonstrates how to render dynamic annotations on a livestream or create alternate video versions with burned-in subtitles by running…
Cloudflare -
CISOs Warned to Govern AI Agents Before Their Access Expands
Artificial intelligence risks and governance are moving from hype to accountability, with Omdia and Gartner outlining how organisations should prepare for 2027. The piece notes that 2026 saw rapid AI deployment across operations, but now boards and security…
Dark Reading -
AI ‘Escapes’ Expose Failures in Guardrails, Not Rogue Intentions
The article argues that the phrase “rogue AI” misleads security thinking by anthropomorphising LLMs and shifting responsibility away from vendors. It stresses that AI agents are nondeterministic software systems, not sentient actors, and that when they…
Dark Reading