Search This Blog

Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Showing posts with label Vulnerabilities and Exploits. Show all posts

Ransomware Group Wiped an Entire VMware Estate in Ten Days Without Exploiting a Single Vulnerability




A ransomware group called BlackNevas dismantled a company's entire virtualization infrastructure over ten days, destroying four ESXi hosts, every virtual machine running on them, two vCenter Server Appliances, and the backup systems meant to recover from exactly this kind of attack. The findings, published by Kaspersky's Global Emergency Response Team, detail how attackers moved through one of the most common enterprise environments without triggering a single exploit alert, because they never used one.

The core finding is uncomfortable for organizations that have treated patching as the main line of defense: every action the attackers took against ESXi ran through legitimate VMware management interfaces. The group obtained valid administrative credentials, used VMware's own tools to push configurations, and triggered ransomware detonation through supported remote access channels.

Ransomware made up 39.4 percent of all incidents Kaspersky's response team handled in 2025. The ESXi segment of that is growing, partly because of a decision made in 2021. When the Babuk ransomware source code leaked, it included a working ESXi version, and at least ten ransomware families have since reused that codebase. BlackNevas is one of them.


How the Attack Unfolded

BlackNevas, first observed in November 2024 and also tracked as a Trigona family variant, operates across Windows, Linux, NAS devices, and ESXi hypervisors. The group runs double extortion, stealing data before encrypting it and threatening public leaks to pressure victims. Known targets include organizations in healthcare, finance, manufacturing, legal services, and telecom across Asia, Europe, and North America.

In this incident, the attackers got in through credential reuse. Valid administrative credentials from non-virtualization systems were used to reach the vCenter Server Appliances, which manage the ESXi hosts. This is the same access path defenders often overlook because they focus on preventing vulnerability exploitation rather than limiting what happens when credentials are already compromised.

Forty-three seconds after logging into the VCSA via SSH as root, the attackers deployed MeshAgent, part of the open-source MeshCentral remote management platform. The entire SSH session lasted two minutes and 26 seconds. Four hours later, they returned through MeshAgent and installed Ligolo-ng, a tunneling tool, as a systemd service disguised under the name `vmware-vdma.service` to blend in with legitimate VMware components.

The next step exploited a feature few organizations think to restrict. VMware's Host Profiles allow administrators to push consistent configurations to ESXi hosts from vCenter. The attackers used this to enable Tech Support Mode SSH across the hosts. Ten minutes and 18 seconds after the configuration was pushed, the TSM-SSH service was running. SSH sessions to ESXi then originated from the vCenter appliance IP, which most environments treat as a trusted internal source.


Taking Out Backups First

Before touching the ESXi layer, the attackers disabled antivirus on the backup infrastructure, stopped backup services, and encrypted the recovery systems. This sequence is deliberate. Ransomware operators have learned that organizations with clean backups can often recover without paying, so the backup environment is now treated as part of the primary attack surface.

Two Veeam vulnerabilities are worth noting here. CVE-2024-40711, which CISA has added to its Known Exploited Vulnerabilities catalog, has been directly linked to ransomware campaigns. CVE-2025-23120 is not on that list yet, but affects domain-joined backup servers and represents a less-patched exposure in many environments.


Detonation and Partial Recovery

The ransomware payload was uploaded to ESXi hosts via SFTP. Roughly three hours before detonation, it ran a self-test lasting about 16 seconds. The final command processed virtual machine storage directories with options to wipe and detach volumes. On one host, all virtual machines stopped within 23 seconds. The ransomware overwrote approximately the first two kilobytes of virtual disk files. Kaspersky notes that a valid backup GPT header may allow partial recovery in some cases, though this is not a reliable path.

A crucial detection problem runs through the entire attack. ESXi writes logs to a RAM disk by default, meaning a reboot erases them. If forensic investigators arrive after a reboot, the evidence of how attackers moved through the environment is gone. Kaspersky recommends configuring remote syslog to preserve ESXi logs outside the virtualization trust domain.


What Organizations Should Change

Kaspersky's recommendations center on treating vCenter as a Tier 0 asset, the same classification given to domain controllers. That means separate administrative identities for virtualization management, phishing-resistant multi-factor authentication, and limiting SSH and ESXi Shell access through Lockdown Mode. Network filtering on the management interface, combined with logs stored outside the virtualization environment, closes the two gaps this attack exploited most effectively.

The practical lesson from this incident is that a patched VMware environment is not a protected one if the credentials managing it are shared with less-secured systems.



Hackers Exploit New N-able N-central Zero-Day Hours After PoC Release

 

Hackers began exploiting a new zero-day vulnerability in N-able N-central within hours of a proof-of-concept exploit becoming public. The flaw, called 4EverAble2Shell, can let unauthenticated attackers execute code with root privileges on vulnerable N-central servers, including systems running the latest available patches. 

The vulnerability was published on GitHub by security researcher Sean Bales, two days after N-able released fixes for several other flaws he had disclosed. 4EverAble2Shell chains an exploit for a recent Apache Axis vulnerability, CVE-2026-90436, with an unspecified flaw affecting N-central deployments that have the Legacy API enabled. The Apache Axis issue has a CVSS score of 9.8, reflecting its critical severity. 

Modus operandi 

The attack chain first exploits the Apache Axis bug to achieve remote code execution. It then attempts to elevate its privileges to root and run a shell script. Researchers said the chain works against N-central servers updated to 2026.4 HF2, build 2026.4.1.26, which had already received fixes for earlier Able2Shell zero-days. This means applying the latest hotfix alone does not fully eliminate the risk. 

N-able responded by restricting several N-central features and integrations, including the Legacy API and Intune integration. The company said it had identified additional indicators of compromise and was working with affected customers to assess and mitigate potential risks. On November 12, Bales also published proof-of-concept code for Agent2Shell, a set of roughly a dozen zero-day flaws that could allow attackers to execute commands on managed Windows agents with System privileges. 

Threat intelligence firm Previdian reported that attackers began using the 4EverAble2Shell proof of concept within hours of publication. WatchTowr also observed probes targeting N-central systems with the Legacy API enabled. Lupovis and Huntback found evidence of successful compromises about five hours after the exploit was released, with attackers deploying an XMRig cryptocurrency miner and using systemd or cron for persistence. 

The attackers named each miner after the victim’s IP address. Lupovis said the attacker’s wallet showed 21 rigs connected to compromised N-central servers in the United States, South Africa, Canada, the United Kingdom, Germany, France, and Sweden. The researchers said their decoy systems did not execute the exploit, confirming that the listed systems were real victims. 

Huntback found that 19 victims had already applied N-able’s latest patches. The firm warned that the miner may be the only observed payload, but the privilege escalation gives attackers root access to a server that manages customer endpoints. Security teams should therefore treat any successful exploitation as a full compromise rather than a minor cleanup task. 

Organisations using N-central should immediately disable the Legacy API where possible, apply vendor guidance, review logs for exploitation indicators, and rotate credentials connected to affected systems. They should also inspect AWS, Azure, and other connected environments for suspicious access, because root-level compromise of an RMM server can expose downstream customer networks.

PostgreSQL’s Largest Ever Security Release Patches 52 Flaws, Including a 22-Year-Old Bug Any Logged-In User Could Weaponize

 


The PostgreSQL Global Development Group shipped security updates on November 12 for all five supported major versions, issuing fixes for 52 vulnerabilities at once: the largest security release the project has published. Thirty-two of the 52 flaws carry a CVSS 3.1 score of 8.8, with no user interaction required, full confidentiality and availability impact, and network reachability through a low-privilege login. The fixed releases are 18.7, 17.12, 16.16, 15.20, and 14.25.

At the center of the batch is CVE-2026-6460, a type confusion flaw in how PostgreSQL handles composite types — data structures built from named fields. When a field's type is changed after values of the original type are already stored, PostgreSQL's binary representation keeps those old bytes intact while the catalog metadata shifts around them. The server then reads a stale value through the new type's output function. That mismatch is enough for a database user who can create objects to run arbitrary code as the OS account running PostgreSQL, no superuser status, no extensions, no file access.

The trigger takes exactly four SQL statements. A user creates a composite type, prepares a query using it, alters the attribute to a different type, then executes the prepared query. The backend selects the wrong output handler, dereferences a null pointer, and crashes. That crash is the entry point.

Security researcher Mehmet Ince found the flaw and published a writeup alongside a demonstration on PostgreSQL 18.6. He traces the root cause to a commit made May 5, 2004, the one that added ALTER TABLE ... ALTER COLUMN ... TYPE. That logic never accounted for prepared statements holding values built against the old type. The code sat there for 22 years. His exploit chain defeats ASLR through a related confusion involving PostgreSQL's unnest() function, then calls system() through a forged object whose methods table points at the system call. He stopped short of publishing that final step. He found 21 variants of the problem in total, and six of the 52 CVEs in this release trace to his work. The release credits 18 researchers for CVE-2026-6460 alone, including PostgreSQL core developers Heikki Linnakangas and Andres Freund, along with the OpenAI Security Research Team and the Anthropic OSS program.

Before going public, Ince spent August coordinating with every major managed provider. He shared working proofs of concept with AWS covering both Aurora and RDS, Microsoft for Azure Database for PostgreSQL, Google for AlloyDB, Supabase, Databricks, and Ubicloud. Google AlloyDB escalated the report to its highest internal priority. Databricks validated the exploits the same day they arrived. Several vendors deployed their own mitigations before the upstream release shipped. Their current patch status is not public, so administrators on managed services should verify the minor version their provider is running.

The release also contains a flaw that skips authentication entirely. CVE-2026-116322, rated 5.3, lets any client restart a PostgreSQL 18 server without logging in. It involves a cancel request and a spinlock held during the authentication timeout window. Two further vulnerabilities sit in the client tools: a crafted archive fed to pg_restore can execute code as whoever invoked it, and a compromised source server can inject commands through the restore script inside pg_dump's tar output to any psql client connecting to it.

The November release permanently closes the PostgreSQL 14 branch. Version 14.25 is its last update. Administrators still on that version are running unsupported software from today. PostgreSQL 15 reaches the same status in November 2027.

Upgrading involves more than swapping binaries. The fix for CVE-2026-6460 changes the server's internal ABI for C-language extensions, which must be rebuilt against the patched version. Extensions compiled for the new release will not load on an unpatched server, so the upgrade and the rebuild must happen together. A new server setting, composite_type_mismatch, enables the type-version checks behind the fix; the project says disabling it carries security risk. Client tools must be updated on every machine running them, not just servers. Clusters managed by Patroni or any custom service unit need a manual restart. Users of incremental backups should take a new full backup before anchoring further incrementals. No workaround exists for the principal vulnerability on servers that cannot be patched, and PostgreSQL has reported no attacks in the wild.

The same acknowledgment list that names Ince, Linnakangas, and Freund also includes AI-assisted tooling. That dynamic cuts both ways. PostgreSQL 19, normally shipped in September or October, was pushed to November 19 partly because AI and LLM tools generated enough new reports to overwhelm the maintainers' review capacity before the feature freeze. Developer Robert Haas noted that LLM-drafted patches still need human review to meet project standards, and consultant Christophe Pettus pointed to the volume of low-quality machine-generated reports as a growing burden. The same tools that surfaced a 22-year-old bug made it harder to ship the next version.


Fortinet Releases Fixes for Nearly 100 Vulnerabilities Across Its Product Lineup

 


Fortinet issued 89 security advisories on November 11, covering close to 100 vulnerabilities across its product lineup, with roughly 17 rated critical and nearly 80 classified as high severity. The release stands as one of the company's largest single-day patch drops, touching nearly every major product in its portfolio.

FortiMail received the heaviest concentration of critical fixes. FortiAuthenticator, FortiOS, FortiManager and FortiAnalyzer followed close behind. FortiSASE and FortiClient also received updates resolving critical issues.

The critical vulnerabilities span a wide range of attack classes. Among those fixed are arbitrary code execution flaws, authentication bypass conditions, account takeover paths, unauthorized access to Virtual Domains (VDOMs), server-side request execution, and a command execution bug that could allow an attacker to run commands with super-admin privileges. FortiMail-specific issues include paths that expose secure email accounts and allow IBE webmail session compromise, along with flaws enabling unauthorized webmail password resets.

High-severity fixes address cross-site scripting, code injection, arbitrary file uploads, denial-of-service conditions, man-in-the-middle attack vectors, privilege escalation, and information disclosure. Fortinet has not attributed any of these vulnerabilities to active exploitation in the wild.


A Notable Standalone Finding

One vulnerability drew particular attention from the security research community. Kevin Joensen, a principal security consultant at Baldur Security, identified and named a FortiClient flaw he called "Fortinet2Shell." The root cause is a CORS misconfiguration that allows attacker-controlled web pages to communicate with a local service running on the target machine.

If a victim visits a malicious website while FortiPAM is enabled, an attacker could achieve arbitrary code execution on that system. The flaw is conditional but the bar for triggering it is low: simply visiting a crafted page is enough. Joensen's research illustrates how client-side software that communicates with local services over HTTP remains a consistent attack surface, one that often gets less scrutiny than server-facing products.


AI in the Testing Pipeline

The scale of this release tracks directly to a shift in how Fortinet approaches internal security testing. Carl Windsor, the company's CISO, confirmed that Fortinet has been running frontier AI models alongside its own on-premises models in vulnerability testing for several months.

The program operates in two phases. In the first, models analyzed firmware without access to source code, mapping how products function and identifying potential attack vectors. Those leads were then validated through penetration testing and red teaming. The black-box phase returned few directly exploitable flaws, but the models showed strong capability in comparing firmware versions across releases and building proof-of-concept exploits for bugs that had already been patched.

The second phase involved source code analysis. Across more than 100 tests spanning 30 product families, roughly 40 percent of findings were genuine vulnerabilities, 30 percent were false positives, and the remaining 30 percent were security improvements that did not rise to the level of exploitability.

Windsor did not frame AI-assisted testing as a silver bullet. He acknowledged that the same tools available to Fortinet's internal teams are available to attackers, and he warned that the gap between public disclosure and active exploitation is getting shorter. "The window between disclosure and exploitation continues to shrink, and we expect it to keep shrinking," he said.


The Urgency Behind Early Disclosure

Fortinet's decision to hold technical details until a fixed release is available reflects a deliberate strategy. Once a patch ships, researchers and threat actors alike can reverse-engineer the changes to identify what was fixed and where. For organizations running unpatched versions, that creates a narrow and shrinking window of practical anonymity.

Windsor noted that internet-facing systems should be prioritized for immediate patching. For organizations that cannot apply fixes right away, interim steps include restricting administrative access to trusted IP ranges, keeping FortiManager off public networks, enforcing strong and unique credentials, and requiring multi-factor authentication for all admin accounts. He was direct that none of these measures replace upgrading the software.

Fortinet's PSIRT page at fortiguard.com/psirt carries the full list of advisories and affected version ranges.

The November batch comes weeks after the company released urgent fixes for FortiOS and FortiManager, some of which were also flagged as AI-assisted discoveries. The cadence suggests Fortinet is now committing to more frequent and larger release cycles as a direct consequence of expanding its AI testing program, a move that will likely increase pressure on enterprise security teams to maintain tighter patch windows across Fortinet deployments.


QNAP Urges Immediate Plex2Shell Update and Remediation

 

QNAP has issued a security advisory concerning a vulnerability chain, collectively referred to as Plex2Shell, affecting Plex Media Server installations on QNAP NAS devices. The issue affects Plex Media Server version 1.43.2 and earlier and enables unauthenticated remote code execution when individual flaws are chained together. QNAP has therefore instructed administrators to update Plex Media Server immediately to version 1.43.3, build 1.43.3.10861 or later. 

Technical scope 

Plex2Shell comprises four vulnerabilities: CVE-2026-96652, CVE-2026-96654, CVE-2026-96655, and CVE-2026-96656. Individually, each issue has a limited security impact, but an attacker can combine them to execute arbitrary code under the privileges assigned to the Plex Media Server process. The vulnerability resides in the third-party Plex application rather than in QTS or QuTS hero. Consequently, QNAP states that its operating systems are not directly affected, and a NAS that does not have Plex Media Server installed does not require remediation.

Affected administrators should navigate to App Center > My Apps > Plex Media Server and select Update. Once the update completes, they must verify in Plex Web under Settings > General that the installed version is 1.43.3.10861 or later. QNAP advises against delaying the update until a later release becomes available. If the update is not visible, administrators should refresh App Center and confirm that the NAS has outbound internet connectivity before attempting the installation again. 

Compromise assessment Updating Plex removes the exploitable condition but does not reverse an intrusion that has already occurred. QNAP states that systems running Plex 1.43.2 or earlier while remotely reachable should be treated as potentially compromised. This exposure applies when Plex Remote Access was enabled, TCP port 32400 was manually forwarded, or the port was exposed through UPnP or myQNAPcloud router configuration. In such cases, QNAP requires a clean reinstall, credential rotation, and a review for indicators of compromise. 

Administrators should inspect running processes for unfamiliar executables launched by Plex, including myawesomeproxy, mylovelyminer, or x12345. They should also examine the Plex Codecs directory for unrecognized .so or .dll files and monitor outbound traffic for connections to proxystream.st. Additional indicators include abnormal CPU utilization when no media is being played and detections produced by QNAP Malware Remover, including signature MR2104. QNAP warns that deleting suspicious files alone is insufficient, since persistence mechanisms may remain; affected systems should follow the full clean-reinstall procedure. 

Secure reinstallation 

For remediation, QNAP recommends preserving the Plex Logs directory if technical support may be required. Administrators should then remove Plex Media Server through App Center, reinstall version 1.43.3 or later, and avoid restoring the previous contents of the Codecs folder. Plex will automatically download required codec components after reinstallation. To reduce recurrence risk, QNAP advises disabling Plex Remote Access when unnecessary, removing port 32400 forwarding rules, disabling UPnP, and using a VPN for legitimate remote access. Administrators should also keep QTS or QuTS hero and all installed applications updated.

SonicWall Patches Exploited VPN Authentication Flaw

 

SonicWall has released security updates addressing a high severity vulnerability in SonicOS that could allow attackers to bypass SAML authentication and its associated multifactor authentication requirements. Tracked as CVE-2026-19868, the flaw affects the SSL VPN component of the operating system used by SonicWall firewalls. The company disclosed the issue alongside six other vulnerabilities in an advisory published on November 6, 2026. Its security team investigated a report indicating potential exploitation, raising concerns for organizations relying on affected appliances to secure remote connections. 

The vulnerability does not allow anyone to enter without credentials. Instead, an attacker must possess valid credentials for an alternative authentication method. Those credentials could then be used to bypass SAML authentication controls, potentially avoiding multifactor authentication enforced through SAML. According to SonicWall, exposure exists only when SAML login is enabled for the SSL VPN. The flaw received a CVSS severity score of 7.1 and was reported by Ryan Harris of The Electronic Office, distinguishing it from an unrestricted authentication bypass affecting every configuration. 

Affected products include Gen7 TZ, NSa, NSsp, and NSv firewalls running SonicOS 7.3.3-7015 or earlier. Gen8 TZ, NSa, and NSv firewalls running SonicOS 8.2.2-8015 or earlier are also vulnerable. SonicWall addressed the issue in SonicOS versions 7.3.4-7013 and 8.2.3-8015, respectively. The company said no workarounds are available and strongly urged customers to upgrade promptly. For administrators, the immediate priority is therefore identifying affected firewall models, checking their installed software versions, and applying the appropriate fixed release rather than relying on an unverified temporary configuration change. 

The United States Cybersecurity and Infrastructure Security Agency added the vulnerability to its Known Exploited Vulnerabilities catalog on November 6. Under Binding Operational Directive 26-04, federal agencies must address it by November 20. However, public information about exploitation remains limited. Neither SonicWall nor CISA disclosed details about the suspected attacks, their targets, or their impact. It also remains unclear whether attackers exploited the weakness before a patch became available. That distinction matters because the available reporting does not establish that this was a zero-day. 

The remaining six vulnerabilities broaden the advisory’s significance, although SonicWall did not identify them as exploited. Two management interface flaws, each rated 8.8, could permit unauthorized actions under certain circumstances without authentication. Three pre-authentication buffer overflows affect the SSL VPN service and TACACS+ authentication client, potentially causing application denial of service. A weak password hashing issue affects virtual appliances but requires highly privileged local access. Together, these findings make the updates relevant beyond the SAML bypass alone, while keeping each vulnerability’s prerequisites and consequences distinct.

Hackers Exploit OpenSSH SSHole Flaw Within Hours

 

A newly disclosed OpenSSH vulnerability, nicknamed SSHole, is already being exploited in the wild. The flaw, tracked as CVE-2026-123158 and carrying a CVSS score of 8.1, lets unauthenticated attackers log in to certain Linux accounts that have no usable password set. Security firm Previdian observed the first exploitation attempt less than nine hours after the vulnerability was published, and confirmed successful root access on an Alpine Linux honeypot. 

The issue affects Linux systems running OpenSSH on the musl C library with Pluggable Authentication Modules (PAM) disabled. Alpine Linux’s default OpenSSH server configuration is among the affected setups. An attacker can send an oversized password—between 257 and 1,024 bytes—to trigger the bug and gain access to accounts configured with an asterisk (*) as their password hash. This means attackers do not need to steal SSH keys or valid credentials to obtain access. 

SSHole stems from password-handling behavior linked to a 2016 patch for CVE-2016-6210, a username-enumeration flaw. Under the specific conditions, OpenSSH mistakenly treats a failed password-hashing operation as successful authentication because it produces the same asterisk value used to mark password-disabled accounts. The SSH daemon then compares that value with the account’s stored password field and allows the login. Importantly, the flaw does not bypass passwords on accounts that have valid password hashes. 

The consequences depend heavily on the privileges of the compromised account. A successful attack grants the attacker the permissions already held by that account, which could include access to deployment credentials, application secrets, and connected infrastructure. From there, intruders may move laterally across systems or disrupt services. The vulnerability does not affect Dropbear, glibc-based Linux distributions, non-Linux systems, or systems that authenticate passwords through PAM.

Administrators should act quickly by installing OpenSSH 10.6p2 or their distribution’s patched package and restarting the sshd service. Alpine also advises reviewing /var/log/messages for “Accepted password for” entries involving accounts that should not accept password logins, including root, Alpine, or key-only users. Administrators should inspect recent users, files, authorized keys, and doas usage; if a suspicious login cannot be ruled out, the system should be treated as compromised.

CISA Warns of Actively Exploited WordPress RCE Flaw

 

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has warned that attackers are actively exploiting a recently patched, high-severity vulnerability in WordPress Core. The flaw, tracked as CVE-2026-119977, can allow unauthenticated remote attackers to execute code on affected servers. CISA added the vulnerability to its Known Exploited Vulnerabilities (KEV) Catalog and ordered federal civilian agencies to address it by Friday under Binding Operational Directive 26-04. The warning highlights how quickly attackers move from public patch information to real-world exploitation. 

The vulnerability is caused by a Time-of-check Time-of-use (TOCTOU) race condition in WordPress’s upgrade process. It was reported by Picus Security researcher Umut Bayram and relates to predictable, web-accessible staging paths used during full-package core upgrades. During the upgrade window, an attacker may be able to interfere with the process and achieve remote code execution. Although exploitation is described as high-complexity, the potential impact is serious because successful attacks do not require prior privileges on the server. 

WordPress released version 7.1.4 on October 28 to fix the issue, urging administrators to update their sites immediately because it was a security release. CISA later confirmed evidence of active exploitation and emphasized that such vulnerabilities are a frequent attack vector for malicious cyber actors, posing significant risks to the federal enterprise. While the federal directive applies only to U.S. federal agencies, CISA also urged private-sector organizations and other network defenders to prioritize patching their WordPress installations. 

The scale of attempted attacks appears substantial. According to WordPress security firm Wordfence, it blocked 13,095 attacks targeting the vulnerability within a 24-hour period. CISA has not publicly disclosed detailed information about the threat actors, targets, or post-exploitation activity. Even so, the volume of blocked attacks suggests that adversaries are scanning broadly for vulnerable WordPress sites rather than focusing on a small number of high-value targets. 

The incident follows earlier WordPress security concerns, including the “wp2shell” vulnerability chain patched in July. That chain combined a REST API batch-route confusion flaw and an SQL injection vulnerability to deploy webshells and take over unpatched servers. Public exploit code appeared shortly after the fix, and attackers later used the technique to install malicious plugins and steal database records. Together, these cases reinforce an essential lesson for site owners: apply WordPress core and plugin updates promptly, restrict administrative access, monitor upgrade activity, and maintain reliable backups to reduce the impact of exploitation.

Attackers Leverage Critical Node-RED Flaw to Gain Unauthorized Access


There is an active exploit of a critical security vulnerability in Node-RED by threat actors to install malicious Linux bots and webshells on vulnerable systems. It is believed that the attacks, detected by cybersecurity firm VulnCheck, could result in compromised instances and industrial environments being exposed to remote command execution and unauthorized access. 

Tracked as CVE-2026-58115, the vulnerability has a CVSS severity rating of 10.0. Node-RED HTTP interface does not contain authentication, allowing remote attackers to access programming functions and execute system commands on the underlying server if unauthenticated remote attackers are not authenticated. 

A browser-based editor enables users to build event-driven applications using Node-RED, an open-source, flow-based programming tool. It is widely used in industries including manufacturing, healthcare, home automation, and industrial control systems. A number of Siemens IoT2050 Advanced devices have been integrated with the platform, including Siemens SIMATIC IoT2050 Advanced devices.

Siemens disclosed the vulnerability on August 11, 2026, after identifying the issue with SIMATIC IoT2050 Advanced devices running Industrial OS with Node-RED installed. In order to take advantage of this vulnerability, attackers could create malicious programming flows and execute arbitrary code with maximum privileges if it is successfully exploited.

A fix for the vulnerability is available in Siemens SIMATIC IoT2050 Advanced version 4.3.4.1 and later. Although security updates have been available, attackers continue to target vulnerable instances. A malicious request containing a complete exploit was captured by one of VulnCheck's monitoring systems on October 30, when it was detected as successful exploitation. This attack deployed a Node-RED flow, triggered its execution automatically, and downloaded a binary from a server controlled by the attacker. 

A malicious execution flow was initiated using Node-RED components, including the inject and execute nodes, without requiring additional interaction from the attacker. There has been no indication of the exact purpose of the downloaded binary, identified as bot-linux-amd64, which was stored under a hidden filename in a memory-backed location and executed in the background. 

Researchers were unable to retrieve the binary for further examination. During the exploitation campaign conducted between October 30 and November 1, repeated attempts were made from the same IP address. The attack vectors used by VulnCheck are 16 attempts to exploit CVE-2026-58115, with attackers delivering a variety of payloads across multiple requests. 

The address used by the attackers originates from Microsoft's AS8075 network, which is located in Brazil. A researcher cautioned that the use of cloud providers' networks does not necessarily identify threat actors, since such infrastructure may include virtual machines rented or compromised by cloud providers. Other attempts were made to install webshells, which gave attackers access to compromised servers and the ability to execute commands remotely, in addition to deploying Linux bots.

In some instances, additional shell scripts and files appeared to be legitimate Linux kernel processes. As opposed to the originally observed bot, which was stored in memory and could be removed when the system was rebooted, the webshells provided a more persistent method of access. In addition to identifying broader scanning activity from the same IP address, VulnCheck also recorded 125 requests targeted at eight different vulnerabilities affecting Apache Solr, Camunda, Gitea, MLflow, Traefik, and Joomla. 

A survey of Node-RED installations indicates opportunistic mass scanning rather than activity concentrating exclusively on Node-RED installations Security experts have expressed concerns about the vulnerability of industrial environments through poorly secured Node-RED deployments. 

Industrial controllers are often integrated with databases, messaging systems and cloud services through this platform, allowing the platform to hold credentials for multiple connected systems. Consequently, a compromised instance could give attackers access beyond the initial server and facilitate movement across networks in the areas of information technology and operational technology. It is also important to note that Node-RED deployments are not always managed by dedicated IT security teams, increasing the risk. 

It is common for engineers and operational staff to configure Node-RED installations directly, thus preventing regular asset inventories and security monitoring. Additionally, scheduled maintenance requirements can delay firmware updates in industrial environments, resulting in vulnerable devices remaining exposed even after patches have been released.

CVE-2026-58115 has been added to VulnCheck's Known Exploited Vulnerabilities (KEV) catalog in response to the observed attacks. When reported, this vulnerability was not listed in CISA's KEV catalog. It is recommended that organizations using Node-RED apply security updates, enable authentication via the adminAuth setting, and restrict external network access to the editor. 

The security team should also examine deployed flows for unanticipated executable nodes, unauthorized HTTP endpoints, and suspicious file modifications. It is equally important to examine system logs and outbound network connections for signs of compromise, as installing patches alone may not remove webshells or other malicious components already present on the system.

Node-RED deployments exposed to industrial environments may pose security risks resulting from the active exploitation of CVE-2026-58115. It is recommended that organizations prioritize patching, enforce authentication, limit access to the editor and investigate for signs of compromise. In light of the possibility of existing webshells remaining after updates have been applied, security checks should accompany remediation efforts as well.

Synology NAS Owners Hit by C3RB3R Ransomware as Old Vulnerability Gets New Life

 



A ransomware strain called C3RB3R is actively hitting Synology network-attached storage devices, locking files and demanding bitcoin payments from victims who, in many cases, were running software patched nearly two years ago.

Synology confirmed the attacks on October 29, pointing to CVE-2024-10441, a critical remote code execution flaw in its DiskStation Manager operating system rated 9.8 out of 10 on the CVSS scale. The vulnerability was found at Pwn2Own 2024 by DEVCORE researchers Pumpkin Chang and Orange Tsai, and Synology patched it by December 2024 under security advisory SA-24:20. The patch has been available ever since. The fact that attackers are still finding success with it tells you something about how many users never applied it.

Once attackers get into a device, C3RB3R goes to work quickly. The ransomware skips 27 system directories and a handful of file types, then recursively encrypts everything else. Each file receives its own 256-bit encryption key generated on the fly, locked with ChaCha8 and then wrapped with RSA-2048, which means there is no way to crack it without the attacker's private key. The malware drops a ransom note called read-me3.txt into every affected folder, containing a six-digit victim ID and a Tor address. Files come out the other side with a .L0CK3D extension. Backups stored on the same NAS are not spared.

The ransom demand starts at 0.036960 bitcoin, which works out to roughly $2,952 at current rates. Victims who miss the five-day deadline see that figure double to 0.073920 BTC, or about $5,904. No free decryptor exists, and Synology has offered none. Affected users who do not have off-site backups are, for practical purposes, out of options.

Reports first surfaced in South Korea around October 28. Victims filed complaints with KISA, the country's internet security agency, but KISA declined to say how many devices were affected, what caused the intrusions, or what it was doing about them. Synology has similarly not disclosed how many users were hit, which device models were targeted, or when it first became aware of the campaign. A Taiwanese NAS research group that compiled damage cases found a common thread: in every instance they reviewed, the device's management interface was reachable directly from the internet, whether through router port forwarding, UPnP, QuickConnect, or a reverse proxy.

The attack timing is not coincidental. On October 25, four days before Synology issued its statement, a public proof-of-concept exploit for CVE-2024-10441 appeared on GitHub. That kind of public PoC typically compresses the window between disclosure and active exploitation to days, sometimes hours. Security researchers at 360 Security Team and Solar Emergency Response Team, who both published analysis on October 29, noted that the campaign appears to have kicked into gear almost immediately after the PoC dropped.

There is, however, a complication in Synology's explanation. The Solar team examined one affected device running DSM 7.4.1-90080 Update 1, a build that does not fit neatly into the SA-24:20 vulnerability timeline. The researchers have not ruled out other entry points, including vulnerable third-party packages, container services, or stolen credentials. No zero-day has been confirmed, but the mismatch has not been fully explained either.

Adding another layer, the 360 Security Team's analysis points to a different set of flaws entirely: CVE-2026-13639 and CVE-2026-13684, both scored 9.8, patched by Synology on September 18, 2026. Under 360's reading, these are "1-day" exploits built on patches that were barely a month old when the attacks began. If 360 is right, devices that only applied the 2024 patch would still be exposed.

Separately, Synology disclosed two more critical vulnerabilities in its PHP infrastructure, CVE-2026-109618 and CVE-2026-109619, affecting packages including Web Station, WordPress, Joomla, and phpMyAdmin. Synology says these flaws are not the C3RB3R entry point, but has advised users to patch them regardless. Security firm Genians found that 96 percent of internal-network devices running PHP were exposed.

Synology's guidance for users is practical if grim for those already affected. The company recommends running a Security Advisor scan, checking Task Scheduler for unfamiliar scripts, and reviewing User and Group settings for administrator accounts that were not created by the device owner. If either check turns up something unexpected, Synology says to contact its support team rather than attempt a manual cleanup. For everyone else, the advice is to update DSM immediately, enable automatic updates, and move backup jobs off-device using Hyper Backup or Snapshot Replication.

The patches for SA-24:20 are in DSM 7.2.1-69057-6, 7.2.2-72806-1, 7.2-64570-4, 7.1.1-42962-7, and 6.2.4-25556-8. Synology PSIRT says it continues to monitor the campaign.



Drupal Commerce Flaw Exposes Online Stores to Fraudulent Order Manipulation

 



Drupal has confirmed active exploitation of a highly critical access bypass vulnerability in its Commerce Core module, warning online retailers that attackers could manipulate orders and circumvent payment processes. The company has urged administrators to install the latest security updates and investigate existing transactions for unauthorized modifications.

Tracked as CVE-2026-118539, the vulnerability affects Drupal Commerce Core, a contributed module that enables websites to sell products and services through shopping carts, checkout systems and payment integrations. It also supports headless commerce architectures, where external applications interact with Drupal through APIs.

According to the original security advisory, certain application routes and entity fields do not have sufficient access restrictions, allowing unauthorized users to bypass established permission checks. This could enable attackers to manipulate commerce-related information, potentially causing direct financial losses for merchants.

Published on October 28 under advisory SA-CONTRIB-2026-254, the vulnerability received a Drupal security risk score of 21 out of 25, placing it in the highly critical category. Its assessment identifies the flaw as exploitable without authentication or special access requirements, regardless of the module's configuration.

Drupal has confirmed that the vulnerability is being actively exploited in the wild. However, its initial advisory did not disclose the attack methodology, affected organizations or the number of compromised websites.


Fraudulent Order Attempts Reported on Live Websites

Additional information published on October 29 by Centarro, the maintainer of Drupal Commerce, provides further insight into the exploitation activity.

According to Centarro, it received two reports from merchants who observed suspicious attempts to place fraudulent orders on their live websites. Following an investigation conducted alongside the Drupal Security Team, the maintainers stated that they had not found evidence of scripted or widespread exploitation beyond the initially reported sites.

The vulnerability affects both traditional, full-stack Drupal Commerce installations and headless deployments, although the potential attack paths differ depending on the website's architecture.

Centarro identified several suspicious transaction patterns that administrators should investigate. These include orders placed without corresponding payments, transactions marked as paid despite having no associated payment records, product prices modified without legitimate promotions, and unauthorized discounts, fees or other adjustments applied to orders.

Some reported attempts involved products listed at unusually low prices, including $0.01, while others used apparently normal prices to make fraudulent transactions less conspicuous.

The maintainers also warned that the vulnerability may have existed in earlier Commerce Core releases for several years. Exploitation was theoretically possible from version 8.x-2.18, released in May 2020, on installations using Drupal's JSON:API module. Other configurations became potentially vulnerable from version 8.x-2.25, released in June 2021.

Consequently, organizations should not restrict their investigation to transactions created shortly before the October security disclosure.


Security Updates and Additional Commerce Vulnerabilities

Drupal has released Commerce Core 3.3.13 and 8.x-2.41 to address CVE-2026-118539. Administrators running the 3.x branch should upgrade to version 3.3.13, while installations using the earlier 8.x-2.x branch should move to 8.x-2.41.

Centarro has also recommended updating the associated Commerce API, Commerce License, Commerce Shipping and Commerce Stripe modules, which received separate security releases. Websites that do not require write operations through JSON:API should consider restricting the interface to read-only access.

The October 28 security releases addressed several additional vulnerabilities across Drupal's contributed projects.

Among them, CVE-2026-118558 affects Commerce Stripe through a cross-site request forgery (CSRF) vulnerability. The module does not include a CSRF token in the return URI used during Stripe Connect authorization. An attacker could exploit this weakness by tricking an administrator into following a malicious link, potentially causing an order to be marked as paid without an actual payment.

Another issue, CVE-2026-118582, affects the CAPTCHA module. The vulnerability results from inadequate invalidation of previously solved CAPTCHA sessions, allowing an already-submitted protected form to be submitted again without completing another challenge. This could weaken protections against automated registrations, comment spam and login brute-force attempts. Drupal's advisory indicates that exploit code exists, although it does not confirm active exploitation.

Drupal also disclosed information disclosure vulnerabilities in Commerce Core and Media Alias Display. The former can expose information about unpublished product variations under certain JSON:API configurations, while the latter involves improperly configured caching that could expose files containing information intended to remain inaccessible.


Earlier Critical Drupal Vulnerabilities

The October disclosures follow another security release on October 14, when Drupal addressed two critical remote code execution vulnerabilities.

CVE-2026-108097 affects the Field UI module in Drupal Core, while CVE-2026-108099 impacts the contributed Entity List module. Although neither component independently exposes the vulnerabilities through its default functionality, certain contributed modules and custom implementations can introduce exploitable pathways. Drupal identified at least one potential attack path that does not require authentication.

Another vulnerability, CVE-2026-9082, involved SQL injection in Drupal Core's database abstraction API, affecting websites using PostgreSQL databases. The flaw could allow unauthenticated attackers to execute specially crafted database queries, potentially resulting in information disclosure, privilege escalation or remote code execution. The vulnerability was added to the US Cybersecurity and Infrastructure Security Agency's Known Exploited Vulnerabilities catalog in May 2026 following confirmed exploitation.

With CVE-2026-118539 now confirmed under active exploitation, administrators should prioritize security updates, examine historical order records and verify that payment confirmations correspond to actual transactions before fulfilling orders. Centarro has also recommended comparing suspicious transactions against website access logs to establish whether they originated through legitimate checkout processes.

ChatGPT Share-Link Flaw Could Expose Private Chats and Hijack Sessions

 

A recently disclosed vulnerability in ChatGPT demonstrated how a maliciously crafted shared conversation link could execute attacker-controlled code within a victim’s logged-in session. The issue was identified by security researchers at Parameter, who demonstrated that the flaw could expose private conversations and allow actions to be performed using the victim’s authenticated account. The vulnerability was reported to OpenAI through its bug bounty program on October 18 and was remediated eight days later. OpenAI classified the issue as P1, its highest severity rating. 

The attack required no software installation, workspace membership, or separate login through the attacker. A victim could receive the malicious ChatGPT share link through email, a messaging platform, a support ticket, or social media. In the most reliable version of the attack, the victim opened the link and clicked the Preview button once. In a later variant, Preview opened automatically, meaning that merely opening the link was sufficient to trigger the attack. 

Once triggered, the malicious code executed within the victim’s ChatGPT session and could interact with ChatGPT’s signed-in APIs. The researchers showed that it could list recent conversations, access a private chat that had never been shared, and transmit part of its contents to an external server under their control. Because the code operated within an authenticated session, it had the potential to perform actions on behalf of the affected user. 

The vulnerability resulted from a combination of weaknesses in ChatGPT’s web architecture. ChatGPT’s Preview feature renders shared HTML in a sandboxed frame on a separate domain, intended to isolate it from the main ChatGPT application. However, an interactive graph in the shared conversation loaded the MathLive mathematics library into the main ChatGPT page. This library retained a message listener that accepted keyboard-related messages from any sender without verifying their origin. The researchers exploited this behavior to send raw HTML from the isolated Preview frame into the trusted main ChatGPT page. 

OpenAI reproduced both the code-execution and private-chat proof-of-concept demonstrations and implemented fixes across two versions of its web frontend. By October 26, both versions rejected the messages on which the attack depended. Parameter published its technical analysis on October 28 after the remediation, and OpenAI awarded a total of $5,000 through its bug bounty program. The researchers also noted that the issue reflected broader weaknesses in message-origin validation, HTML handling, and sandboxing, rather than a problem that could be resolved simply by blocking third-party scripts such as YouTube.

Millions of Routers Running Broadcom Wi-Fi Chips Have an Unpatched Flaw That Hands Attackers Root Access Over the Air

 




A security researcher has demonstrated working exploits against a class of Wi-Fi gateways running Broadcom chips, showing that an attacker within radio range can gain root-level code execution on the device without knowing the Wi-Fi password or completing any kind of pairing process. As of this writing, no patch has been deployed to affected routers, and no CVE identifier has been assigned.

Quentin Kaiser published his findings on October 16, 2026, after a disclosure process that effectively went nowhere. The vulnerability sits inside Broadcom's own implementation of Wi-Fi Protected Setup, the handshake protocol designed to simplify connecting new devices to a network. Kaiser found that the code mishandles the reassembly of fragmented WPS messages. By sending carefully crafted fragments, an attacker can corrupt the device's memory and redirect execution, all through the air, from a position outside the building.


Two Working Exploits, Two Different Outcomes

Kaiser demonstrated the attack against two Belgian ISP-supplied gateways. On an Orange Belgium device, an Askey TCG3000, the exploit read the Wi-Fi pre-shared key directly from stored settings and transmitted it back to the attacker, effectively handing over the network password. On a Proximus gateway, the Sagemcom BBOX3V+, the exploit did something more severe: it opened a reverse shell transported entirely over Wi-Fi frames, without requiring the attacker to have an external server or any prior foothold on the network.

Both attacks required no credentials and no interaction from anyone inside the network. Distance helps the attacker, not the defender.


The Code Is Old, and So Is Broadcom's Awareness

The vulnerable code traces back to at least 2010. Broadcom issued a partial fix in 2018 and simultaneously told downstream customers to migrate to the hostapd Wi-Fi software package instead of relying on Broadcom's own WPS implementation. That guidance went unheeded across a large portion of the device ecosystem, and the underlying flaw was never fully resolved.

Kaiser says Broadcom was informed about the issue again through his disclosure process and declined to assign a CVE or comment publicly. CERT/CC, which typically coordinates vulnerability disclosures between researchers and vendors, closed his case the day it was opened, citing the fact that Broadcom holds its own CVE numbering authority.

The result is a vulnerability without an identifier, without a patch, and without any official acknowledgment from the company whose code is responsible.


Which Hardware Is Affected

Kaiser sorted the affected chips into two groups. Older eCos-based chips, the BCM3380, BCM3383, and BCM33834, cannot receive a practical fix given their architecture. He described these as a "forever day," meaning a vulnerability that will remain open for the functional life of the hardware. Newer Linux-based chips, including the BCM3390, BCM65xx, and BCM68xx series, are technically fixable if device makers choose to replace the WPS software stack.

Among confirmed deployments in Belgium, Proximus, Orange Belgium, and VOO gateways are affected. Kaiser was unable to test Telenet devices but considers them potentially in scope. His broader analysis, which involved automated firmware scanning, identified vulnerable code across Cisco small-business access points, Huawei gear, and ARRIS routers, among others. He estimates the total number of affected devices runs into the millions globally, though some entries in his published list are flagged as "likely vulnerable" rather than confirmed.


Detection Tools, but No Mitigation

For organizations trying to assess their exposure, Kaiser published YARA detection rules and a scanning script on GitHub at github.com/qkaiser/broadcom-wps. Both are intended as triage tools to identify firmware containing the vulnerable code. He notes that patched devices may still trigger the YARA rules, so a positive result requires additional verification rather than automatic remediation.

For devices that fall into the "forever day" category, the practical answer is replacement.


Why Proximity Attacks Are Getting More Attention

The threat model here matters beyond the technical details. Kaiser noted that this vulnerability is particularly useful to attackers who can position themselves in physical range of a target rather than approaching through the internet. That scenario is no longer theoretical.

In November 2024, Volexity documented how a Russian state-sponsored group, tracked as GruesomeLarch and linked to APT28, compromised organizations near a high-value target in Washington, D.C., and then used dual-homed systems inside those neighboring buildings to connect to the actual target's corporate Wi-Fi. The attacker never needed to be in the same room or even the same building. They just needed a wireless network with weak enough controls to reach from the street.

A flaw that lets anyone within radio range become root on the router, without a password, without a handshake, fits neatly into that playbook. The attack surface is wherever the signal reaches.



AI Spam Surge Forces Google to Pause Open Source Bug Bounty Program


OSS VRP (Open Source Software Vulnerability Rewards Program) has temporarily been suspended after a sharp increase in automated reports that were found to be valid. Since October 1, security teams and open-source maintainers have been facing an increasing number of low-quality vulnerability reports generated via artificial intelligence. 


The pause, which took effect on October 1, is in response to increasing volumes of low-quality vulnerability reports. Google announced the OSS VRP in August 2022 as a means of rewarding researchers who identifies and responsibly discloses security flaws in open-source software maintained by the company. 

The program covers projects such as Golang, Angular, Bazel, Protocol Buffers and Fuchsia, along with selected third-party dependencies. Security concerns regarding GitHub Actions, application configurations, repository settings, and access control rules are also covered by this program. There was initially a range of rewards available from $100 to $31,337 under the program, with particular emphasis placed on vulnerabilities that could potentially pose significant risks for software suppliers. 

As a result of the company's wider vulnerability rewards program, millions of dollars have since been awarded to researchers, making the OSS VRP an important means of identifying security vulnerabilities in widely used open-source projects. An increase in automated submissions was responsible for the current suspension, according to the company, with the majority failing to identify valid security issues. 

Although the use of artificial intelligence-assisted tools has made the generation of vulnerability reports at scale easier, the resulting volume may also include incorrect findings, duplicate claims, and reports concerning vulnerabilities that do not exist. OSS VRP is currently being reviewed by Google to address the issue and determine how the program should handle the growing number of automated submissions. 

A further update is anticipated in the first quarter of 2027. Additionally, the company clarifies that the change does not affect outstanding reports or product vulnerabilities submitted prior to October 1. The impact of the AI-driven reporting surge extends beyond Google's program as well. 


It has not been the company's first time experiencing a sharp increase in low-quality vulnerability submissions that have been generated by automated tools. This raises concerns about the time security teams will need to spend validating reports that do not identify genuine vulnerabilities. The Google Patch Rewards Program continues to offer incentives to researchers for submitting high-impact open-source security patches that qualify for rewards of up to $15,000. 

Google Cloud's Cloud Vulnerability Reward Program provides a means of reporting security vulnerabilities affecting open-source repositories related to Google Cloud products. Google's decision follows similar developments elsewhere in the security industry. In January, the curl project maintainer terminated its HackerOne bug bounty program in response to a significant number of low-quality, artificial intelligence-generated vulnerability reports. 

As part of its Intigriti bug bounty program, Intel also removed financial rewards in September, though the company did not provide a publicly stated reason for the change. As the use of artificial intelligence tools increases in speed, potential vulnerabilities are identified and reported more rapidly, while the review process remains the responsibility of security researchers and maintainers. 

Earlier this year, Microsoft warned that AI-assisted vulnerability discovery could increase the number and scale of security discoveries, potentially increasing the operational demands on security teams. Google has not yet confirmed that the Open Source OSS VRP will be permanently discontinued. The company is reviewing the program and anticipates making changes in the first quarter of 2027. 

For now, the temporary suspension reflects a growing challenge for bug bounty programs, namely, how to handle a large volume of automated reports without allowing invalid findings to overwhelm genuine security issues. 

The decision of Google highlights the difficulty of vulnerability reward programs as AI-assisted security research increases submissions. It will become increasingly important for these programs to distinguish genuine findings from automated and inaccurate reports in order for them to be effective.

Google Introduces Gemini 4 Argon With Guardrail-Free Access for Defenders

A new frontier artificial intelligence model, Gemini 4 Argon, has been introduced by Google through its Fairwind Program for initial distribution to trusted cybersecurity defenders. In addition to internal security teams using this model, the company expects wider access as it collects feedback from early users. 

As a software engineering, enterprise knowledge work, and cybersecurity operations solution, Argon is designed to handle complex software engineering and knowledge management tasks. A model developed by Google will be able to identify, validate and patch critical vulnerabilities independently in security environments, thereby expanding the use of artificial intelligence for vulnerability research and remediation. 

Argon will be available to trusted defenders and the company's own teams without cyber-specific guardrails, according to the company. As part of this approach, vetted security professionals will be given full access to the model's capabilities when investigating and addressing threats. In September, Fairwind, a limited access AI security tool for governments, Google Cloud customers and cybersecurity partners, launched.

A significant finding has already been made as a result of its early deployment, Wiz, which is using Argon as part of its Scan for Good initiative, reported that it identified a previously unknown critical vulnerability in healthcare software used by hospitals worldwide. The vulnerability may expose sensitive personal information, although Google has not disclosed the name of the affected software or whether the issue has been resolved. 

Google also reports significantly improved vulnerability detection performance compared with Gemini 3.8 Flash Cyber. A security test conducted by Argon on complex codebases identified security weaknesses, while a test conducted by Wiz on live web applications demonstrated improvements in attack surface discovery, vulnerability identification, and proof-of-concept generation. 

A phased approach is being taken by Google to the wider release, with the model currently restricted to internal teams and vetted defenders. Moreover, the company is participating in the U.S. government's voluntary pre-release process and will refine its safeguards after receiving feedback from early testers in order to broaden the availability to developers, enterprises, and individuals. 

Argon will be designed to reject requests attempting to support cyber or chemical, biological, radiological, and nuclear attacks as part of its broader rollout, while also preserving the support of legitimate dual-purpose research as part of its broader rollout. Additionally, Google is monitoring the model's internal activity for signs of misuse. Indirect prompt injection is also being investigated. 

In Google's opinion, Argon is protected against attempts to manipulate it through malicious instructions or external content. The Fairwind program provides another layer of control around access by monitoring the model’s reasoning and actions, and stopping execution when behavior goes beyond the intended task. 

Organizations participating in the program have been vetted and their use has been restricted to authorized defense activities such as threat simulation, reverse engineering, and malware analysis for research or security purposes. Partners are not permitted to share or distribute access to the model. Google has not provided a date of general availability yet. 

Upon initial deployment of Argon Defender, API customers and Google AI Ultra subscribers should have access, although the broader deployment of Argon will be dependent on the results of ongoing safety and security evaluations.

MCP Python SDK Flaw Exposes OAuth Credentials

 

A high-severity security vulnerability in the official Model Context Protocol (MCP) Python SDK could allow malicious MCP servers to steal OAuth credentials from artificial intelligence applications. Tracked as GHSA-qx49-fqc8-xw99, the flaw has a CVSS score of 7.5 and affects HTTP-based MCP clients that use OAuth authentication while connecting to servers that are not fully trusted. The issue does not affect local studio clients, MCP servers, or applications that provide their own authentication tokens and headers. 

The vulnerability is linked to the SDK’s OAuth discovery process. During authentication, an MCP client asks a server to identify the authorization server responsible for issuing tokens. A malicious server could return a 404 response for the standard OAuth discovery endpoint, forcing the SDK to use a fallback method. On this unsafe path, the SDK failed to properly validate the authorization server’s issuer, allowing the attacker to redirect the authentication flow to an infrastructure controlled by them. 

As a result, attackers could capture an OAuth client secret, authorization code and PKCE proof key. These credentials may enable the attacker to obtain valid access tokens from the legitimate identity provider, potentially gaining the same permissions as the affected application. Depending on the configured OAuth scopes, the impact could include access to cloud services, internal APIs, databases, deployment systems and other connected resources. Long-lived client secrets and refresh tokens could also support continued access or account takeover. 

The affected releases include MCP Python SDK versions 1.9.1 through 1.29.1 and versions 2.0.0 through 2.1.1. The maintainers fixed the issue in version 1.30.0 for the 1.x branch and version 2.2.0 for the 2.x branch. However, upgrading alone may not be sufficient for deployments using ClientCredentialsOAuthProvider or PrivateKeyJWTOAuthProvider; developers must explicitly configure the expected issuer= value. Users of the deprecated 1.x RFC7523OAuthClientProvider should migrate to another supported provider. 

Organizations should immediately identify affected clients, upgrade the SDK and remove stored OAuth client registrations created by older releases. If a vulnerable client connected to an untrusted MCP server, administrators should rotate client secrets, revoke potentially exposed tokens and review authentication logs for suspicious activity. Teams unable to upgrade should restrict connections to MCP servers they completely control and trust. The advisory highlights the security risks created when AI agents connect external tools to privileged enterprise systems, making strict server verification and issuer validation essential safeguards.

Hackers Breach Polish Medical Software Firm Qbusoft, Expose Patient Data in Second Healthcare Attack in Weeks


 


A cyberattack on Polish healthcare software company Qbusoft has left patient records from its Medyc platform potentially in the hands of attackers, coming just weeks after a separate, larger breach hit another Polish medical software provider and rattled the country's entire health data infrastructure.

The attacker exploited an SQL injection vulnerability in Medyc's application interface during late August, according to a breach notification published last week by the Addiction and Psychiatric Treatment Center in Inowrocław, one of the healthcare facilities running the platform. SQL injection is one of the oldest and best-documented attack techniques in security research, allowing an attacker to manipulate a web application into pulling data directly from its database. Despite decades of awareness about the flaw, it remains a recurring entry point in healthcare system compromises.

Qbusoft confirmed on Friday that the attackers obtained names, national identification numbers, home addresses, phone numbers and email addresses. In Poland, the national identification number, called a PESEL, functions similarly to a Social Security number in the United States and is a standard credential for identity verification across government services, banking and healthcare. Its theft puts affected patients at real risk of identity fraud.

The company said it had not confirmed the theft of clinical records. But the Inowrocław center told patients that Qbusoft found evidence the attacker ran scripts specifically targeting database tables containing medical information, making it "highly likely" that medical records were also pulled. The data in scope for that facility included hospital treatment records and discharge summaries from patients treated at its Day Treatment Unit for Addiction Treatment between July 2024 and August 2026.

The intrusion occurred on August 22-23 and went undetected until the night of September 8-9, a gap of more than two weeks. By that point, the attacker had already transferred an encrypted archive of the database outside Qbusoft's systems. Some fields, including names and PESEL numbers, had been encrypted in the database. Qbusoft nonetheless advised the affected center to assume the attackers could decrypt that information without difficulty, given the specifics of how the protection was implemented.

Qbusoft patched the vulnerability on the day the breach was detected, restricted database access permissions, rotated passwords and technical credentials, and introduced additional monitoring. The company has not publicly commented on the incident through any official statement.

That silence drew a sharp response from Digital Affairs Minister Krzysztof Gawkowski, who said the Central Bureau for Combating Cybercrime had opened an investigation and criticized Qbusoft for failing to notify CERT Polska or the national incident response team for the healthcare sector before authorities reached out. "Hiding attacks by companies is the biggest mistake, as it always puts citizens at risk," Gawkowski said. Poland's data protection authority separately announced that its president had ordered a formal audit of Qbusoft.

Medyc, which has operated as a cloud-based platform since 2014, is used across Polish medical practices and clinics for electronic medical records, patient scheduling, electronic prescriptions, referrals, sick notes, telemedicine and administrative billing. In a public notice, the company warned that its infrastructure had faced repeated attack attempts since the incident and that users might see temporary slowdowns or restricted access to certain modules.


The Same Attacker?

Polish cybersecurity publication Zaufana Trzecia Strona reported that a person or group using the alias "fingerprint" contacted the outlet claiming responsibility for the Medyc attack. The publication had previously linked that alias to the MyDr breach, a separate incident involving another Polish healthcare software vendor. Polish broadcaster RMF FM also reported that the same attackers behind MyDr were likely responsible for the Medyc intrusion, though Polish authorities have not formally attributed the attack to any individual or group.

The alleged attacker claimed to have obtained records on 5 million patients and 8 million private photographs, some of which Zaufana Trzecia Strona said may depict patients in sensitive medical settings. Neither figure has been independently confirmed, and the stolen data has not been made public. The actor reportedly framed the operations as an effort to expose weak security rather than profit from the data.

The MyDr breach, confirmed in August, potentially affected close to 19 million people across more than 12,000 healthcare facilities, involving over 2 terabytes of stolen data including names, PESEL numbers, prescription histories, diagnoses and appointment records. Poland has roughly 36.5 million residents, meaning the MyDr incident alone touched the records of nearly half the country's population. The Inowrocław treatment center caught up in the Medyc breach was also among the organizations affected by MyDr.

Gawkowski said Polish authorities had observed a surge in criminal activity targeting healthcare organizations in recent weeks and were preparing new regulations in response, including mandatory security certification for healthcare technology companies and tighter controls on how private vendors handle medical data. Poland recorded a 144 percent year-on-year rise in reported cybersecurity incidents in 2025. The consecutive breaches of Medyc and MyDr, both software vendors connecting thousands of clinics to national health infrastructure, have made clear that the weakest link in Poland's health data chain is not the government platform but the private companies sitting in front of it.