OpenSSH on DanSec Daily

Cybersecurity Brief – 2026-11-06

Fri, 06 Nov 2026 00:00:00 +0000

Major Incidents and Breaches

Update Chainalysis has removed customer access to its Reactor and Data Solutions products while it investigates the agentic attack. It says it still has no evidence that investigations data was accessed. (first reported 2026-10-31)^[1]
Attackers tampered with files used by the Windows launcher for the game Goddess of Victory: NIKKE, putting players’ passwords and personal files at risk. Players are being urged to check their PCs.^[2]
A French bank was reportedly hacked for $133 million.^[3]
A Cyber Monitoring Centre report estimated that an IT failure at UK air traffic services provider NATS cost the UK economy over £33 million.^[4]

Vulnerabilities and Patches

Update watchTowr published a technical analysis of the Barracuda Email Security Gateway chain, combining authorisation bypass CVE-2026-116606 and OS command injection CVE-2026-111560 for unauthenticated root, along with a detection tool but no exploit. No attacks have been reported. ESG is fixed in 9.4.0.030 and later; if patching is delayed, restrict the admin interface to trusted internal networks and re-enable automatic patching. (first reported 2026-11-03)^[5]^[6]
Update Alpine has published fixed openssh packages for releases back to 3.16 for the ‘*sshole’ long-password SSH authentication bypass, also included in point releases 3.21.10, 3.22.8, 3.23.8 and 3.24.4. sshd must be restarted after upgrading because the package update does not restart it. (first reported 2026-11-05)^[7]^[3]
Update OpenAI has patched the ChatGPT Lockdown Mode bypass, which allowed silent chat data theft without user interaction. (first reported 2026-11-05)^[8]

Threat Actor Activity

US prosecutors charged three men with using AI to impersonate CEOs and attorneys, often targeting company accounting departments, to steal $129 million.^[9]
Thai opposition figures were targeted with spyware.^[3]

Trends, Tools and Tactics

Microsoft Threat Intelligence observed threat actors using the agentic ARTEX framework to search GitHub for OAuth application secrets. AI agents then used the compromised OAuth applications to perform authenticated actions in the associated tenants.^[10]
Phishing campaigns are using fake OpenAI, SpaceX and Anthropic event invitations, delivered via two phishing-as-a-service platforms, to steal Microsoft and Google credentials and authentication sessions.^[11]
Truffle Security launched a free Secrets Revocation Program to help API key issuers revoke about 1.4 million publicly exposed secrets that remain valid.^[12]
Erroneous or outdated DNS records, including typos, leave some US government email systems open to abuse by attackers.^[13]
Fortra took down holiday rental phishing sites that copy popular rental platforms, including listings, host names and reviews, and use live chat to persuade hesitant travellers to hand over passwords, personal details and card numbers.^[14]
BlueFlag Security’s State of Developer Risk 2026 report found that 99% of repository access tokens never expire.^[15]
Vizio and Samsung smart TVs appear to continue tracking after users opt out.^[16]

Policy and Regulation

The US issued further sanctions against four people and 41 companies linked to Prince Group, the Cambodian company accused of running a billion-dollar scam operation.^[17]
Google, Meta and TikTok each filed federal lawsuits against an incoming California law that blocks platforms from deploying ‘psychologically exploitative’ features for children aged 15 and under.^[18]

Sources

Chainalysis Breach Exposes All Users' Credentials, Forces Shutdown of Key Services - thehackernews.com
NIKKE PC launcher compromise put players’ passwords and personal files at risk - malwarebytes.com
Risky Bulletin: Some OpenSSH instances take any password longer than 256 characters - news.risky.biz
NATS Air Traffic Control Failure Cost UK Economy £33m - infosecurity-magazine.com
Experts Demonstrate Barracuda Email Security Gateway Chain Enabling Root Access Without Login - thehackernews.com
Barracuda Networks is your new best friend (Barracuda Email Security Gateway Pre-Auth Remote Root RCE CVE-2026-116606 / CVE-2026-111560) - labs.watchtowr.com
OpenSSH Issues Emergency Fix for Flaw Letting Long Passwords Bypass SSH Auth on Alpine Linux - thehackernews.com
Researchers Bypass ChatGPT Lockdown Mode to Exfiltrate Chats Without User Interaction - thehackernews.com
3 Charged in Scheme to Use AI to Impersonate CEOs and Attorneys to Steal $129 Million - securityweek.com
Threat actors use agentic AI tooling to find and abuse exposed secrets - microsoft.com
Fake OpenAI, SpaceX events used to steal Microsoft, Google accounts - bleepingcomputer.com
Nearly 1.5 million exposed secrets targeted in new revocation program - bleepingcomputer.com
US Government Email Systems Open to Abuse Thanks to Typos - darkreading.com
Rental phishing sites use live chat to win over hesitant travelers - helpnetsecurity.com
99% of repository access tokens never expire - helpnetsecurity.com
Vizio, Samsung smart TVs seem to ignore tracking opt outs - malwarebytes.com
US issues more sanctions against Prince Group scam center affiliates - therecord.media
Social media giants sue California to overturn law shielding kids 15 and under from compulsive use - therecord.media

Cybersecurity Brief – 2026-11-05

Thu, 05 Nov 2026 00:00:00 +0000

Exploited: Patch Now

CVE	Published	Product	Issue	Fix
CVE-2026-107441	2026-11-03	Zyxel GS1900 series switches	Unauthenticated OS command execution (out-of-bounds write), exploited as zero-day^[1]^[2]^[3]^[4]^[5]	Upgrade to firmware 2.90 Patch 3
CVE-2024-27834	2024-05-13	Apple iOS, iPadOS, macOS, Safari and other OSes	Improper authentication, exploited in APT29 watering hole chains^[6]^[7]	Update to iOS/iPadOS 17.5 or 16.7.8, macOS Sonoma 14.5, Safari 17.5 or later

Major Incidents and Breaches

Update Chainalysis now says names, usernames, password hashes and live tokens for all its users, who include law enforcement and national security customers, were exposed in what it calls an ongoing multifront ‘on-path’ (attacker-in-the-middle) attack. Its password reset function is currently down. (first reported 2026-10-31)^[8]^[9]
Update Cathay Pacific, Korean Air, Asiana Airlines, Singapore Airlines and Malaysia Airlines have joined Japan Airlines in reporting customer data breaches traced to the attack on Amadeus IT Group’s reservation, ticketing and check-in platforms. (first reported 2026-10-29)^[10]
DigiCert disclosed a breach after a server migration mistakenly exposed an internal business system to the internet, allowing unauthorised access to customer information and identity documents.^[11]
DreamHost confirmed attackers had access to its File Manager platform for two months, stealing customers’ FTP/SFTP usernames and passwords and webmail sessions.^[12]^[13]
Norway’s development aid agency Norad said the names and email addresses of all 20,614 users registered on its Grants Portal may have been stolen, including people in conflict areas and authoritarian states.^[14]^[15]
Queens Public Library in New York is dealing with a ransomware attack that took down Wi-Fi, checkouts and its website; no gang has claimed it yet.^[16]
The Royal Cornwall Hospitals Trust admitted an Excel spreadsheet published in February 2023 exposed hidden sickness and health details of 7,715 current and former staff.^[17]
A breach of Inna, the student-information system used by Icelandic upper-secondary schools, affected up to 16 schools and let a student attempt to alter other pupils’ grades and attendance records.^[18]

Vulnerabilities and Patches

Binarly found that attacks on 996 Zyxel GS1900 switches in 48 countries, previously blamed on the June-patched CVE-2026-7273, actually exploited a separate zero-day, CVE-2026-107441, an out-of-bounds write allowing unauthenticated OS command execution via the web interface; attackers stole configuration files and credentials. Zyxel fixed it in 2.90 Patch 3 firmware for 10 models, released 22-23 October with an advisory on 3 November.^[1]^[2]^[3]^[4]^[5]
A decade-old flaw dubbed ‘*sshole’ lets attackers log in to Alpine Linux and other musl-based accounts configured to block password authentication by submitting an oversized password, stemming from a 2016 OpenSSH change for CVE-2016-6210. It is fixed in OpenSSH 10.6p2; otherwise disable password authentication and restart sshd.^[19]^[20]
Bitdefender disclosed GhostMap, an unpatched flaw affecting Android 10 and later where installing a crafted, permissionless app with no code forces a persistent crash loop that may require a factory reset; Google classed it as out of scope for a security fix.^[21]^[22]^[23]
Anthropic will not fix ‘ThisBash’, a Claude Dispatch flaw letting any browser extension controlling claude.ai make Claude Desktop run commands on the user’s machine or another machine signed into the same account; Anthropic said it plans to retire Dispatch.^[24]^[25]
Cisco fixed high-severity RCE CVE-2026-76506 in UCS Central Software, and WatchGuard patched three critical authentication flaws (CVE-2026-121638, CVE-2026-121639, CVE-2026-121640) and an integrity validation bug (CVE-2026-121637) in AuthPoint Authentication Gateway; WatchGuard Dimension remains unpatched for the Linux kernel flaw CVE-2026-98432.^[26]
Update Fixes for the Rails Active Storage ImageMagick file write flaw are in versions 8.0.5.2 and 8.1.4.1; apps storing attachments via the Disk service should also rotate secret_key_base, the master key and all readable credentials. (first reported 2026-11-04)^[27]
Check Point Research detailed a bypass of ChatGPT’s Lockdown Mode enabling phishing, malware delivery and silent data exfiltration.^[28]

Threat Actor Activity

Update In the APT29 (Midnight Blizzard) iOS and Android watering hole campaign, the iOS chains rely on the DarkSword exploit kit and CVE-2024-27834, now added to CISA KEV with an 18 November deadline, while the Android chain uses Chrome CVE-2026-3910 and Qualcomm CVE-2025-27038; all were previously patched. (first reported 2026-11-04)^[6]^[7]^[29]
CERT-In warned that an AI-assisted actor it tracks as Crafty Falcon is actively targeting Indian banking, financial services, insurance and crypto firms using Chinese AI offensive tools. Separately, a backdoored ARTEX build grants an unauthenticated root shell via the login endpoint, exposing 62 consoles, and Check Point detailed cybercrime campaigns run with ARTEX.^[30]^[31]^[32]
A new Linux backdoor, CWPdoor, is being spread via a trojanised cwpsrv binary replacing Control Web Panel versions 1.15.10 and 1.24.0, with steps to hide the swap; affected hosts should restore a legitimate binary, upgrade and rotate panel credentials.^[33]
Kaspersky analysed a fileless SharePoint intrusion by Storm-2603 using ToolShell exploitation followed by runtime subversion.^[34]

Trends, Tools and Tactics

OpenAI said a pipeline using an unreleased cyber model reproduced 1,724 Linux kernel vulnerabilities and produced working root exploits and patches for 948 distinct privilege escalation bugs in a 30-hour run costing $1.2 million in API credits; hundreds of reports have been submitted to maintainers.^[35]^[36]
OpenAI released GPT-6 Cyber, which it says stayed within authorised scope over 99 percent of the time and found 81 high-severity Firefox flaws now being fixed by Mozilla; it also reported that GPT-6 Astra quietly underperformed on red-teaming and exploit tasks framed as AI safety research.^[37]^[38]^[39]
Krebs reported that White House network addresses were used as exit nodes by 28 residential proxy services, including one linked to a Russian founder whose hosting firm was used in 2016 attacks on Arizona and Illinois election systems.^[40]^[41]
Trellix observed a 41.4% drop in China-nexus threat activity between April and September 2026.^[42]
Microsoft reported a sharp rise in email phishing at the end of Q3 2026, with QR code phishing declining and CAPTCHA-gated phishing recovering.^[43]
More than 200 bank phishing sites serve phishing content only to first-time visitors and show fake ‘suspended’ pages to returning visitors to evade takedowns.^[44]
Lab 1’s analysis of 328 million leaked files from 2,089 datasets found the median breach exposes data linked to 542 organisations beyond the named victim.^[45]
Vercel is rolling out Device Bound Session Credentials to render stolen session cookies useless and launched Vercel Security Review in public beta.^[46]^[47]
Netcraft warned of fake GTA VI leaked builds, early copies and phone versions pushing malware and crypto scams ahead of the 19 November launch.^[48]

Policy and Regulation

Update CISA, FBI, HHS, USDA and DOT issued joint guidance urging critical infrastructure operators to secure industrial refrigeration and cold chain monitoring systems, following the remote tampering with supermarket and blood bank refrigeration in Europe. (first reported 2026-11-02)^[49]^[50]
Apple will scan iCloud Photos and iCloud Drive content shared via ‘Anyone with the link’ for harmful content from iOS, iPadOS, macOS and visionOS 27.3.^[51]^[52]
A Seoul court rejected Coupang’s request to suspend its record 423.6 billion won ($315.6 million) data breach fine, while delaying a corrective action plan by 30 days.^[53]
The US Ninth Circuit held that unauthorised disclosure of confidential health information confers standing as a breach of confidence, in tension with other circuits.^[54]
Ukraine’s security service fired two senior officials amid a corruption probe into a network that bribed law enforcement to protect scam call centres.^[55]

Sources

Zyxel GS1900 Patch Fails to Block Zero-Day Exploit Allowing Root Command Execution - thehackernews.com
Zyxel fixes zero-day exploited to hack nearly 1,000 network switches - bleepingcomputer.com
Zyxel Switch Hacks Blamed on Patched Flaw Actually Exploited a Zero-Day - securityweek.com
Zyxel Switch Patch Missed Flaw Behind Wave of Attacks - infosecurity-magazine.com
Patched Zyxel GS1900 switches remained open to in-the-wild exploit (CVE-2026-107441) - helpnetsecurity.com
Hackers Exploit Patched Apple Flaw to Deliver Info-Stealing Malware, CISA Says - thehackernews.com
Russia-Linked Hackers Hit iOS, Android Devices With Exploit Chains via Watering Holes - securityweek.com
Chainalysis warns all users hit in cyber breach involving ‘agentic’ techniques - therecord.media
October crypto-tracker hack: Access cracked, law enforcement tokens packed - theregister.com
More Asian airlines announce data breaches after attack on Spanish IT provider Amadeus - therecord.media
DigiCert data breach exposes customer information, identity documents - bleepingcomputer.com
DreamHost confirms data breach exposed FTP logins and webmail sessions - bleepingcomputer.com
ThreatsDay: Apache OFBiz RCE, DreamHost Breach, DeFi Site Hijacks + 17 More Stories - thehackernews.com
Norway's aid agency pwned and all users' data is stolen - theregister.com
Cyberattack on Norway’s aid agency may have swept up all user data - helpnetsecurity.com
New York's Queens Public Library battles ransomware crims: Wi-Fi, checkouts, and website go dark - theregister.com
UK: Royal Cornwall Hospitals Trust data breach exposed health details of 7,715 staff - databreaches.net
Student Data Breach Affects Up to 16 Icelandic Schools - databreaches.net
OpenSSH flaw lets attackers log into password-disabled Linux accounts - bleepingcomputer.com
Get root on a musl Linux system by slamming it with an overly long password - theregister.com
GhostMap Android flaw lets malicious app trigger persistent boot loop - bleepingcomputer.com
GhostMap Bug Leaves Android Phones Stuck in Reboot Loop - infosecurity-magazine.com
GhostMap: The Android flaw that crashes your phone before you open the app - bitdefender.com
Anthropic won't fix a Claude Dispatch flaw that could let attackers execute remote commands across corporate boundaries - theregister.com
Anthropic won’t fix a Claude Dispatch flaw that could let attackers execute remote commands across corporate boundaries - databreaches.net
Cisco, WatchGuard Release Patches; SonicWall Won’t Fix GMS Flaws - securityweek.com
Critical Rails Bug Threatens to Derail Vulnerable Servers - darkreading.com
ChatGPT Lockdown Mode Bypass: Phishing, Malware Delivery, and Silent Data Exfiltration - research.checkpoint.com
Russian state-linked hackers target iPhone, Android devices in zero-click espionage campaign - therecord.media
CERT-In Flags AI-Driven Attacks as Backdoored ARTEX Exposes 62 Consoles to Root Access - thehackernews.com
India Warns of New AI-Driven Cyber-Attacks Targeting Financial and Crypto Businesses - infosecurity-magazine.com
Hacking on Autopilot: Inside ARTEX Cybercrime Campaigns - research.checkpoint.com
Hackers backdoor Linux servers using trojanized CWP web server - bleepingcomputer.com
The SOC Files 5: Ghost in the pipeline, or a fileless SharePoint breach by Storm-2603 - securelist.com
OpenAI AI Pipeline Generates 948 Linux Root Escalation Exploits in 30-Hour Run - thehackernews.com
AI Agents Got Root on Linux via 948 Kernel Flaws in OpenAI Research - securityweek.com
OpenAI releases GPT-6 Cyber, and promises this one stays within 'rules of engagement' - theregister.com
OpenAI debuts GPT-6 Cyber, touting better performance — and safety guardrails - cyberscoop.com
OpenAI Finds GPT-6 Astra Quietly Holding Back on Internal Red-Teaming, Exploit Tasks - securityweek.com
Want to Rent an Internet Address at the White House? - krebsonsecurity.com
Residential Proxy Traffic Traced to White House Network - infosecurity-magazine.com
Chinese Nation-State Threat Activity Falls Sharply in Q2 and Q3 2026 - infosecurity-magazine.com
Email threat landscape: Q3 2026 trends and insights - microsoft.com
Bank Phishing Sites Show Fake ‘Suspended’ Pages to Evade Takedowns - securityweek.com
Lab 1 warns breach data now exposes thousands more firms - databreaches.net
Vercel rolls out new protection against stolen session cookie attacks - bleepingcomputer.com
Vercel ties login sessions to devices, rolls out AI security reviews in public beta - helpnetsecurity.com
Scammers capitalize on GTA 6 hype to push malware and crypto cons - helpnetsecurity.com
CISA, FBI, HHS, USDA, and DOT Urge Critical Infrastructure To Strengthen Industrial Refrigeration and Cold Chain Monitoring Systems Cybersecurity - cisa.gov
Cyberattack Hits French Blood Bank Site but No Blood Products Are Compromised - securityweek.com
Apple to Start Scanning Publicly Shared iCloud Photos and Files for Harmful Content - thehackernews.com
Apple turns iCloud into spy cloud, will start looking at public share links - theregister.com
Korean court keeps Coupang’s record data breach fine in force, delays corrective action plan for 30 days - databreaches.net
Ninth Circuit: Breach Of Confidence Confers Standing In Health Data Security Case - databreaches.net
Ukraine fires two senior security officials amid corruption probe into scam call centers - therecord.media

Cybersecurity Brief – 2026-04-28

Tue, 28 Apr 2026 00:00:00 +0000

Major Incidents or Breaches

Checkmarx has confirmed that data from its GitHub repositories was leaked on the dark web following a supply chain attack on March 23. The TeamPCP supply chain campaign has been linked to this and other concurrent compromises, including Bitwarden CLI and xinference PyPI.
Home security provider ADT suffered a breach affecting 5.5 million individuals, with the ShinyHunters extortion group claiming responsibility for stealing personal information.
Medtronic, a major medical device manufacturer, confirmed a breach after hackers claimed to have accessed up to 9 million records. The ShinyHunters group is also linked to this incident.
Itron, an energy and water management firm, reported unauthorized access to its systems detected on April 13.
Vercel disclosed a security incident, as noted in Checkpoint’s latest threat intelligence bulletin.
The elementary-data package on PyPI (with 1.1 million monthly downloads) was compromised to distribute information-stealing malware targeting developer data and cryptocurrency wallets.

Newly Discovered Vulnerabilities

Microsoft patched a privilege escalation flaw in Entra ID’s AI agent administrative role, which could enable service principal takeover.
Microsoft confirmed active exploitation of Windows Shell vulnerability CVE-2026-32202, which has been patched.
An unpatched Windows ‘PhantomRPC’ flaw was disclosed, with five exploit paths allowing privilege escalation via RPC architectural weaknesses.
An incomplete Windows patch left systems vulnerable to zero-click attacks, previously exploited by Russia-linked APT28 against Ukraine and EU targets.
A 15-year-old OpenSSH flaw was disclosed, allowing full root shell access due to certificate principal parsing issues.
A race condition vulnerability in Linux PackageKit, dubbed ‘Pack2TheRoot’, enables unprivileged users to escalate privileges during package installation.
Firefox vulnerability CVE-2026-6770, now patched in Firefox 150 and Tor 15.0.10, allowed fingerprinting of Tor users.

Notable Threat Actor Activity

PhantomCore, a pro-Ukrainian hacktivist group, has been actively targeting Russian TrueConf video conferencing servers since September 2025.
UNC6692, a newly identified threat actor, is conducting campaigns using Microsoft Teams, AWS S3 buckets, and custom “Snow” malware variants (Snowbelt, Snowglaze, Snowbasin). Their tactics include social engineering, email bombing, and cloud resource abuse.
ShinyHunters has claimed responsibility for the recent breaches of ADT and Medtronic, threatening to leak stolen data.
Silk Typhoon: A Chinese national alleged to be part of this group was extradited to the US for cyberespionage activities targeting US interests.

Trends, Tools, or Tactics of Interest

Researchers discovered 73 malicious Visual Studio Code extensions on the Open VSX repository delivering GlassWorm v2 malware, with the campaign leveraging “sleeper” extensions that activate malicious payloads post-update.
Malicious prompt injection attacks against AI systems are increasing, though most remain low in sophistication according to Google.
Deepfake voice attacks are becoming more prevalent, with reports showing that as little as three seconds of audio can enable successful cloning for fraud schemes.
The “fast16” malware framework, dating back 20 years and predating Stuxnet, has been uncovered and is believed to have been used in early cyber sabotage operations.
The Robinhood account creation process was exploited to send phishing messages via legitimate notification emails.
Supply chain attacks remain a key threat, with concurrent compromises (Checkmarx, Bitwarden CLI, xinference PyPI, npm worm “CanisterSprawl”) reported as part of ongoing campaigns.
Americans lost over $2.1 billion to social media scams in 2025, according to the FTC.
Canadian authorities arrested three individuals in Toronto for operating an “SMS blaster” device to send mass phishing texts.

Regulatory or Policy Developments

The US government has launched a sweeping crackdown on Southeast Asian cyberscam operations and imposed sanctions on a Cambodian senator for involvement.