Roundcube Webmail has released security updates for its 1.6 LTS and 1.7 branches, fixing 12 vulnerabilities that could expose users and servers to cross-site scripting, email header injection, cross-user data access, remote-content bypasses, and server-side request forgery attacks.

The new releases, Roundcube 1.6.19 and 1.7.4, address flaws in how the open-source webmail platform processes email content, HTML, Cascading Style Sheets, attachment metadata, contact groups, and remote URLs. Administrators running production deployments of Roundcube 1.6.x or 1.7.x are urged to update as soon as possible.

One of the most serious issues fixed is a zero-click stored cross-site scripting vulnerability involving the injection of TNEF MIME tags into attachment URLs.

TNEF, or Transport Neutral Encapsulation Format, is commonly associated with Microsoft Outlook attachments. An attacker could potentially send a specially crafted email that triggers malicious script execution when the victim views the message, without requiring the user to click a link or open an attachment.

The updates also fix another XSS issue in Roundcube’s HTML editor when handling text/enriched email content. Cross-site scripting weaknesses can allow attackers to execute JavaScript in a victim’s webmail session, creating opportunities to steal session tokens, alter mailbox settings, read messages, or perform actions as the logged-in user.

Several fixes address email header injection risks. These bugs affected the subject field, recipient display name, and an identity’s organization field.

Header injection vulnerabilities can be abused to manipulate email metadata or insert unexpected mail headers if malicious input is not correctly sanitized.

Roundcube also patched a cross-user access issue in SQL-based address books. The flaw involved adding or removing members from contact groups.

It could allow one user to modify another user’s group associations under certain conditions. This type of issue can compromise contact privacy and the integrity of address book data in shared or hosted Roundcube environments.

Remote-content protections received multiple fixes, addressing CSS declaration smuggling, HTML body background property injection, CSS-escape bypasses in FuncIRI attributes, and SVG SMIL source animation techniques that could bypass remote-content blocking.

Roundcube Webmail Patches 12 Security Flaws

The updates further fix an is_local_url() validation bypass involving fully qualified domain names with a trailing dot in stylesheet URLs. Attackers could exploit differences in URL parsing to make an external resource appear local and bypass intended restrictions.

A server-side request forgery bypass was also resolved in the Roundcube CSS proxy. The weakness involved hexadecimal IPv6-mapped IPv4 addresses, which could potentially help an attacker bypass address validation and force the server to request internal or restricted network resources.

Roundcube said full technical details are available in the release notes for versions 1.6.19 and 1.7.4. The project strongly recommends that all organizations operating affected Roundcube installations apply the updates promptly.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Roundcube Webmail Patches 12 Security Flaws, Including Zero-Click XSS and SSRF Bypass appeared first on Cyber Security News.

Security researchers have uncovered a significant vulnerability chain in Telerik UI for ASP.NET AJAX, allowing unauthenticated attackers to execute remote code in vulnerable enterprise web applications.

The issue primarily affects Telerik’s RadAsyncUpload component, a widely used file-upload control in ASP.NET WebForms applications.

Progress Software has indicated that the flaw impacts versions from 2010.1.309 to 2026.2.519. The vulnerability was addressed in version 2026.2.708, released as part of the 2026 Q2 SP1 update.

The vulnerability chain includes four distinct flaws: CVE-2026-13181, CVE-2026-13182, CVE-2026-13183, and CVE-2026-13184. While these vulnerabilities are serious, their exploitation requires specific conditions and cannot be applied universally to all default Telerik deployments.

At the heart of the issue is CVE-2026-13182, a padding oracle vulnerability within RadAsyncUpload’s handling of encrypted client states. Telerik employs AES-CBC encryption to safeguard configuration data exchanged between the server and the user’s browser.

A padding oracle occurs when the application provides different error responses for invalid encrypted data. In this case, malformed data results in a distinct error compared to valid padding with invalid JSON content.

Telerik Flaw Chain

This discrepancy enables an attacker to submit modified ciphertext repeatedly, gathering information on how the application decrypts it, ultimately allowing them to recover sensitive data and forge modified encrypted values without needing the encryption key.

Even when the ASP.NET customErrors feature is enabled, researchers noted that exploitation remains possible, albeit more challenging and time-consuming through timing analysis.

Exploiting this oracle, researchers manipulated Telerik’s serializedConfiguration data, which governs settings within the upload control. This enabled attackers to alter the AllowedFileExtensions field, permitting DLL files to be uploaded.

The attack utilized a CBC forgery technique, introducing a “sacrificial” encrypted block within a JSON string. This method preserved necessary configuration from legitimate page loads, including session controls, while inserting malicious entries in the configuration.

The second critical vulnerability, CVE-2026-13181, pertains to the management of upload metadata, where Telerik resolves the .NET type name supplied via the AsyncUploadTypeName value without a proper allowlist.

If a server-side FileUploaded handler reads the UploadResult property, Telerik deserializes corrupt data into the designated type. This behavior can be exploited together with the System.Configuration.Install.AssemblyInstaller gadget, enabling the application to load an uploaded mixed-mode DLL from a temporary directory, executing native code via its DllMain entry point.

The proof-of-concept demonstrated execution of a web shell within the IIS worker process, while an in-memory variant could run commands without writing any files to disk.

To successfully exploit this vulnerability chain, attackers must access a page containing a RadAsyncUpload control with an active server-side FileUploaded event handler that reads UploadResult.

Additionally, an explicit, non-default Telerik.AsyncUpload.ConfigurationEncryptionKey must be configured for the exploitation path to function, which is recommended as a security measure.

Organizations utilizing Telerik UI for ASP.NET AJAX are urged to upgrade immediately to version 2026.2.708 or later. It is critical for administrators to identify pages using RadAsyncUpload and review their upload event handlers to monitor for potential exploitation.

Vigilance against suspicious IIS activity is also necessary, with particular attention to instances of w3wp.exe unexpectedly spawning cmd.exe, the appearance of DLL files in temporary folders, and unexpected .aspx files in web roots.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Telerik Flaw Chain Lets Unauthenticated Attackers Turn Padding Oracle Into Remote Code Execution appeared first on Cyber Security News.

N-able has released N-central 2026.3 Hotfix 4 to fix CVE-2026-86218. This critical vulnerability could allow an unauthenticated attacker to execute code remotely on an exposed N-central server.

The update, identified as build 2026.3.1.14, was issued for on-premises N-central deployments. N-able urged self-hosted customers to install the hotfix immediately, warning that systems left unpatched remain at risk even though the company has not confirmed exploitation in production environments.

CVE-2026-86218 is a pre-authenticated remote code execution vulnerability. This means an attacker may be able to trigger the flaw without first logging in or providing valid user credentials. If successfully exploited, the issue could allow an attacker to run commands on the N-central server.

N-central is used by managed service providers and IT teams to monitor, manage, automate, and secure customer systems. Because the platform can have broad access across endpoints, networks, credentials, and administrative tools, a compromise of the central management server could create serious downstream risks.

Attackers who gain control of an N-central server could potentially use that access to deploy malicious software, alter monitoring settings, steal stored information, create unauthorized accounts, or move further into managed customer environments.

N-able Released Hotfix

The exact technical details and attack vector for CVE-2026-86218 have not been publicly disclosed. N-able said a third party responsibly reported the flaw through its security disclosure program. The vendor stated that it currently has no confirmation of active exploitation.

However, organizations should not treat the lack of known attacks as a reason to delay patching. Public patch releases can help threat actors identify vulnerable systems and develop exploit attempts.

The new release replaces N-central 2026.3 Hotfix 3, build 2026.3.1.13. Customers running versions 2025.4, 2026.1, 2026.2, 2026.3, 2026.3.1 Hotfix 1, or 2026.3.1 Hotfix 2 can upgrade directly to build 2026.3.1.14. Organizations using older releases should first move to a supported upgrade version and then apply the latest hotfix.

N-able confirmed that hosted N-central customers, also known as NCOD users, do not need to take any action because the patches have already been applied to their environments. The urgent action applies to organizations operating their own self-hosted N-central infrastructure.

The company also said administrators do not need to upgrade N-central agents specifically to address CVE-2026-86218. However, it recommended keeping agents up to date with the latest available version as a general security practice.

Security teams should identify all self-hosted N-central instances, confirm their installed build number, and schedule the update to 2026.3.1.14 as soon as possible.

Administrators should also review server access logs, administrator account activity, remote command execution records, and unusual configuration changes for signs of suspicious behavior before and after patching.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post N-able Released Hotfix for RCE Vulnerability Affecting Platform appeared first on Cyber Security News.

Plex has issued an urgent security update for Plex Media Server and Plex Desktop, asking users to install the latest releases as soon as possible.

The update addresses multiple undisclosed security issues affecting Plex Media Server versions 1.43.2 and earlier, as well as affected Plex Desktop installations. Plex released Plex Media Server version 1.43.3 and Plex Desktop version 1.115.0 to resolve the flaws.

The company has not yet published technical details about the vulnerabilities, including their severity, attack requirements, affected components, or potential impact. Plex said it has requested CVE identifiers and will provide more information after they are published.

The lack of public vulnerability details means administrators should treat the update as a priority, especially where Plex servers are exposed to the internet, use remote-access features, or run on systems holding large personal media libraries.

A vulnerable media server could potentially create opportunities for unauthorized access, data exposure, service disruption, or further compromise, depending on the nature of the flaws.

Plex Fixes Multiple Security Flaws

Plex Media Server is widely deployed on Windows, macOS, Linux, network-attached storage devices, Docker environments, and NVIDIA Shield devices.

This broad platform support makes patch management important because update availability may differ by operating system and hardware vendor.

Windows and macOS users with automatic updates enabled should confirm that their Plex Media Server is running version 1.43.3 or later.

Users who do not receive the update automatically can download the newest server package from Plex’s official Media Server downloads page. Plex also recommends that Desktop users move to Plex Desktop version 1.115.0.

Linux administrators should download the appropriate installation file for their distribution. Ubuntu and Debian-based systems use the .deb package, while Fedora and CentOS-based deployments use the .rpm package.

Administrators should verify the downloaded filename before running installation commands and should restart or validate the service after deployment. NAS users may experience delays because vendor-managed application stores do not always publish new Plex packages immediately.

Plex said affected users can manually download the correct package for their NAS model and install it via the device’s web-based app management interface. The company pointed users toward vendor-specific instructions for QNAP, TerraMaster, Western Digital, Netgear, and Synology systems.

Docker users should follow Plex’s official container deployment guidance and ensure they pull and deploy an image containing the fixed Plex Media Server release.

Organizations running Plex in containers should also review image tags, container restart policies, exposed ports, remote access settings, and reverse-proxy configurations.

Administrators should not rely solely on an update notification. They should open the Plex server dashboard or package manager and confirm the installed version after patching.

Security teams may also want to review server logs for unexpected login attempts, unusual remote connections, new administrator sessions, or unexplained configuration changes while waiting for further CVE and technical disclosure details from Plex.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Plex Urges Users to Update Media Server Immediately to Fix Multiple Security Flaws appeared first on Cyber Security News.

TP-Link has disclosed two security vulnerabilities in its Archer AX55 v4 router that could let attackers on the local network crash a service, steal administrator credentials, and potentially execute remote code on affected devices.

The flaws, tracked as CVE-2026-18167 and CVE-2026-18330, affect the EasyMesh and web login modules in Archer AX55 hardware version V4.

TP-Link released firmware version 1.2.1 Build 20260527 to address both issues. The company published its advisory on September 3, 2026.

The most serious issue, CVE-2026-18167, is a stack-based buffer overflow in the router’s EasyMesh component. It has a CVSS v4 score of 7.7 and is rated High severity.

EasyMesh connects compatible networking devices into a single mesh Wi-Fi network. According to TP-Link, the vulnerability becomes exploitable when Mesh mode is enabled on the Archer AX55 v4.

Multiple TP-Link Archer Vulnerabilities

An attacker connected to the target’s local network could send specially crafted input to the EasyMesh service, known as the easymesh daemon. The malicious input could force the service to crash.

In some cases, the flaw could also allow the attacker to run code on the router. Remote code execution on a router is particularly dangerous because the device sits between local systems and the internet.

Attackers who compromise a router may attempt to monitor network traffic, alter DNS settings, redirect users to malicious websites, scan connected devices, or use the router as a foothold to attack the wider network.

TP-Link said successful exploitation could have a high impact on the confidentiality, integrity, and availability of the affected router. However, the attack requires local network access, and Mesh mode must be enabled.

The second vulnerability, CVE-2026-18330, affects the Archer AX55 v4 web login module. The flaw is caused by a hardcoded shared RSA-1024 private key embedded in the product.

A local attacker who captures an HTTP-based administrator login session could use the known private key to decrypt the administrator password. TP-Link also noted that a weak AES session key reduces the effort required to compromise the login session’s confidentiality.

The issue received a CVSS v4 score of 6.1 and is rated Medium severity. Although it does not directly provide code execution, stolen router administrator credentials could give an attacker control over key configuration settings.

The weakness highlights the risks of using HTTP for administrative access. Unencrypted HTTP sessions can expose sensitive login data to attackers on the same network, especially on insecure or shared Wi-Fi networks.

The vulnerabilities affect TP-Link Archer AX55 routers with hardware version V4. The fixed firmware version is 1.2.1 Build 20260527. TP-Link strongly recommends that owners update their devices as soon as possible through the official Archer AX55 V4 firmware download page.

Users should also turn off Mesh mode when not needed, avoid managing the router over HTTP, use a strong, unique administrator password, and ensure that router management access is not exposed to untrusted networks.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Multiple TP-Link Archer Vulnerabilities Allow Attackers to Execute Remote Code appeared first on Cyber Security News.

Google has released an emergency Chrome security update that fixes a critical zero-day vulnerability already being exploited in real-world attacks.

The flaw, tracked as CVE-2026-85046, affects the V8 JavaScript and WebAssembly engine used by Chrome to process web content. The company confirmed that it is aware of an exploit for the vulnerability existing in the wild.

While Google did not disclose details about the attacks, affected targets, or the threat actors behind the activity, the active exploitation notice makes immediate patching important for all Chrome desktop users.

The security update moves Chrome Stable to version 152.0.7977.82/.83 for Windows and macOS. Linux users receive version 152.0.7977.82. Google said the update will roll out gradually over the coming days and weeks.

Chrome 0-Day Vulnerability Exploited

CVE-2026-85046 is described as a high-severity type confusion vulnerability in V8. Type confusion bugs occur when software incorrectly treats a piece of data as one type when it is actually another type.

In a browser engine, this kind of memory-handling error can be dangerous. An attacker may be able to create specially crafted JavaScript or web content that causes Chrome to handle memory unexpectedly.

Depending on the exploit chain, this could lead to browser crashes, data exposure, or the execution of attacker-controlled code within the browser process.

A victim may only need to visit a malicious or compromised website for an exploit attempt to begin. Attackers can also deliver exploit links via phishing emails, malicious advertisements, social media messages, or compromised legitimate sites.

The vulnerability was reported by security researcher Salvatore Gulizia, also known as Serotav, on August 4, 2026. Google awarded a $1,000 bug bounty for the report.

The Chrome update contains 12 security fixes in total. Several of the patched issues are rated high severity and affect important browser components, including V8, WebGL, Network, DevTools, Skia, CacheStorage, Compositing, and CrashReporting.

Among the notable fixes are a race condition in V8, an out-of-bounds write vulnerability in WebGL, use-after-free flaws in Compositing, DevTools, and Skia, and a type confusion issue in Compositing.

Google is restricting access to some vulnerability details until most users have installed the update. This approach is intended to reduce the chance that attackers can quickly reverse-engineer the fixes and build additional exploit code before organizations and individuals patch their browsers.

Chrome users should update immediately by opening the browser menu, selecting Help, and then choosing About Google Chrome. Chrome will check for the latest version and download the update when available.

Users should relaunch the browser after installation to ensure the security fixes are applied. Enterprise administrators should verify that managed endpoints are updated to Chrome 152.0.7977.82 or later.

Security teams should also monitor endpoint telemetry and web-filtering logs for suspicious browser activity, especially phishing campaigns that use newly registered domains or deceptive update prompts.

Because CVE-2026-85046 is actively exploited, delaying the update leaves systems exposed to a known browser attack path.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Critical Chrome 0-Day Vulnerability Actively Exploited in the Wild appeared first on Cyber Security News.

A critical vulnerability in Sangoma Switchvox is being actively exploited, affecting the enterprise VoIP platform used to manage business phone systems, voicemail, call forwarding, monitoring, and analytics.

The flaw, tracked as CVE-2026-9586, enables unauthenticated attackers to execute commands remotely on vulnerable systems without needing valid credentials.

Horizon3.ai researchers observed valid exploitation attempts against internet-exposed Switchvox devices on August 30, 2026, with attackers attempting to deploy reverse shells for remote command-line access to compromised VoIP servers.

CVE-2026-9586 is an unauthenticated SQL injection vulnerability affecting Sangoma Switchvox SMB Edition 8.3, build 104997, and earlier releases. The issue has a CVSS severity score of 9.3 and can lead to remote code execution.

The vulnerable component is an unauthenticated HTTP endpoint, /pa, that processes XML messages for supported IP phones. These messages can be used to notify phone systems about events such as incoming and outgoing calls.

Hackers Exploiting Sangoma Switchvox VoIP RCE Flaw

Horizon3 researchers found that Switchvox extracts the PhoneIP value from an XML request and directly adds it to a PostgreSQL database query.

The application does not properly sanitize or parameterize the user-controlled input. This allows a remote attacker to inject malicious SQL commands through a specially crafted request.

Because the database query is executed with elevated PostgreSQL permissions, an attacker could run operating system commands on the Switchvox server.

This could allow them to access database contents, alter user data, create or elevate administrator accounts, steal authentication material, and establish persistent remote access.

Horizon3 and Defused Cyber deployed internet-facing honeypots to monitor for exploitation attempts. On August 30, researchers detected attacks from the IP address 176.65.148.184 across multiple honeypots in quick succession.

The observed activity included an attempt to launch a reverse shell using Netcat. Attackers then used a Base64-encoded command to collect information on active processes running on the vulnerable Switchvox appliance.

The results were prepared for transmission to an attacker-controlled server, suggesting that the attackers were conducting post-exploitation reconnaissance after gaining access.

Horizon3 warned that the speed and scale of the attempts indicate that internet-exposed Switchvox systems are likely being broadly scanned and targeted. Researchers identified approximately 4,000 Switchvox devices exposed online, with most located in the United States.

Horizon3 reported 12 Switchvox vulnerabilities to Sangoma in April 2026. Sangoma released Switchvox version 8.4.0.2 on July 14, 2026, which addresses CVE-2026-9586 and other reported issues.

Organizations should immediately upgrade Switchvox to version 8.4.0.2 or a later supported release. Administrators should also review /var/log/switchvox/db-quirks.log for suspicious SQL statements, investigate connections involving 176.65.148.184, and restrict access to the /pa endpoint to trusted phone networks where possible.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Hackers Actively Exploiting Sangoma Switchvox VoIP Platform RCE Flaw in Attacks appeared first on Cyber Security News.

A high-severity vulnerability in the All-in-One WP Migration and Backup plugin could allow unauthenticated attackers to take over vulnerable WordPress sites. The flaw, tracked as CVE-2026-19949, affects more than 5 million active installations and has been fixed in version 7.110.

The issue was reported to Wordfence on August 14, 2026, by security researcher Jack Taylor through the Wordfence Bug Bounty Program. Taylor received a $5,761 bounty for discovering the vulnerability, which received a CVSS score of 8.8.

All-in-One WP Migration and Backup is widely used to export, import, restore, and migrate WordPress sites. It creates .wpress archive files containing website files and database data.

The vulnerable versions, up to and including 7.109, contain an unauthenticated second-order SQL injection flaw in the archive restore process.

Unlike a typical SQL injection attack, the malicious SQL code is not executed immediately. An attacker can first place a specially crafted payload in a WordPress site through the core trackback feature. Trackbacks can be submitted without logging in when a public post accepts pings.

WordPress All-in-One WP Migration Plugin Flaw

The attacker submits malicious trackback data with a carefully prepared blog name and URL. WordPress stores that data in the comments table. At this stage, the payload remains inactive and appears to be ordinary comment-related data.

The attack becomes dangerous when a site administrator exports the website using the plugin and later restores it. During restoration, All-in-One WP Migration rewrites URLs and database table prefixes in SQL statements before importing them.

According to a Wordfence report, a flaw in the plugin’s regular expression handling of backslashes and quoted strings can cause the stored payload to escape its intended SQL string boundary.

As a result, attacker-controlled content can become executable SQL during the database restore process. The injected SQL can retrieve the plugin’s ai1wm_secret_key, a secret value used to protect the plugin’s unauthenticated import action.

The attacker can then leak this secret key into an approved comment and retrieve it through the site’s public WordPress REST API. With the secret key, the attacker may access the plugin’s import process and upload a malicious .wpress archive.

A crafted archive could include a malicious must-use WordPress plugin. Since must-use plugins load automatically, the malicious code can execute when a visitor or administrator opens a page.

This can grant the attacker remote code execution on the server and enable a full site compromise, including deploying a webshell, stealing data, or installing additional malware.

The exploit requires an administrator to perform an export followed by an import after the malicious trackbacks have been planted. Although this adds an interaction requirement, backup and restore operations are routine tasks for many WordPress administrators.

Wordfence deployed a firewall rule for Premium, Care, and Response users on August 16, 2026. Free Wordfence users are scheduled to receive the protection on September 15, 2026. ServMask acknowledged the report on August 17 and released version 7.110 on August 20.

Website owners using All-in-One WP Migration and Backup should update immediately, turn off unnecessary trackbacks, review suspicious comments, and verify that no unauthorized plugins or administrator accounts have been added.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post WordPress Plugin Flaw Exposes 5 Million Sites to SQL Injection Attacks appeared first on Cyber Security News.

Researchers used Claude AI to help port a pre-authentication remote code execution exploit to a WAGO programmable logic controller, demonstrating how AI can assist with low-level operational technology exploitation.

The experiment achieved arbitrary ARM shellcode execution on a WAGO 750-831 PLC without valid credentials. However, it required major human involvement, expensive API usage, and eventually bricked the test device.

The exploit targeted CVE-2021-31886, a buffer overflow affecting the Nucleus FTP server used in several embedded products. The flaw exists because the FTP service fails to validate the length of a username submitted through the USER command.

A specially crafted oversized username can overwrite memory and redirect program execution. Forescout researchers focused on a WAGO 750-831 PLC running firmware V01.04.16. Researchers already had a working exploit for the related WAGO 750-852 model.

Claude AI Builds Pre-Auth RCE Exploit for WAGO PLC

They used Claude to identify the new device’s target-specific memory addresses, function offsets, callback locations, and shellcode storage regions.

Unlike conventional web or application exploitation, the PLC test involved a closed-source embedded device with no access to a debugger.

Claude Code used firmware analysis tools, Ghidra, generated Python scripts, and network testing utilities to inspect the firmware and interact with the live controller.

However, the AI initially followed incorrect code paths, generated invalid exploit attempts, and required researchers to provide disassembly details and steer the analysis.

The first successful result was a crash, confirming that the overflow could be triggered. Achieving reliable RCE proved much harder.

Forescout researchers found that normal FTP command processing erased the attacker-controlled buffer after the USER command was handled, preventing the injected ARM shellcode from surviving long enough to execute.

Claude identified that the exploit could preserve the payload by changing the FTP command sequence. Instead of using the original USER followed by QUIT flow, the final approach used a USER followed by CWD sequence and omitted the CRLF command terminator.

This prevented the normal completion path from clearing the memory buffer. After solving the buffer-preservation problem, Claude quickly generated working payloads. One payload made the PLC send ICMP echo requests to an attacker-controlled system.

At the same time, another transmitted a UDP packet containing the text “PWNED.” The exploit required network access to FTP port 21. However, it did not require authentication because the vulnerable USER command is processed before login.

The final RCE development stage consumed $535.74 in API usage. It lasted 8 hours and 32 minutes across several research sessions; most of the effort involved reverse-engineering the FTP processing chain and understanding why the shellcode was overwritten.

Researchers then attempted to extend the proof of concept into a command-and-control implant. During memory probing and payload testing, one AI-generated payload wrote to flash-mapped memory, permanently bricking the PLC.

The incident illustrates the safety risks of autonomous or semi-autonomous testing against cyber-physical devices. The findings show that AI can help adapt embedded exploits across similar OT targets, although specialist expertise remains essential.

Organizations should restrict unnecessary exposure to FTP and remote management, monitor PLCs for crashes and unexpected outbound traffic, and treat difficult-to-exploit OT flaws as potentially more significant as AI reduces the effort required to develop exploits.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Claude AI Builds Pre-Auth RCE Exploit for WAGO PLC to Execute ARM Shellcode Without Credentials appeared first on Cyber Security News.

FreeRDP released version 3.31.0, addressing 22 security flaws and multiple bugs in its open-source Remote Desktop Protocol implementation, and urges users and distributors to update promptly.

FreeRDP is widely used by Linux systems, thin clients, remote-access tools, and enterprise applications to connect to Windows Remote Desktop Services.

Because it processes network data from remote servers and supports features such as graphics, smart cards, USB redirection, clipboard sharing, and authentication, memory-handling mistakes can create serious security exposure.

The 3.31.0 release includes fixes for 22 GitHub Security Advisories, covering issues reported through the project’s security process. The advisory identifiers include GHSA-c5gr-hmqp-pwj4, GHSA-h5w2-q35j-443h, GHSA-m85m-3qxv-63h5, and 19 others.

While the release notes do not provide public technical details for every flaw, the vendor’s “update ASAP” warning shows that maintainers consider the addressed issues significant.

FreeRDP Fixes 22 Security Flaws

Several changes in the release point to security-sensitive code paths. FreeRDP fixed bounds checking in the AVC444v2 YUV decoder, which processes remote desktop graphics data.

It also corrected parsing and length-validation problems in dynamic virtual channels, Remote Desktop Gateway tunnel responses, clipboard format lists, smart-card data, USB redirection, and device-redirection components.

The update further resolves use-after-free conditions involving the printer driver singleton and the reallocation of aligned memory. Use-after-free bugs occur when software continues to access memory after it has been released.

Depending on the affected code path and surrounding protections, such flaws can lead to application crashes, information disclosure, or possibly remote code execution. Authentication and cryptographic components also received attention.

The release improves NTLM and SSPI memory handling, adds checks before accessing signature buffers, fixes SPNEGO mechanism fallback behavior, and improves error handling when BIO or SSL object creation fails.

These changes are important because FreeRDP often handles authentication exchanges and encrypted connections to remote systems.

In addition to security fixes, version 3.31.0 brings performance improvements. The project said an optimized YUV decoder should deliver faster client-side graphics for AVC and H.264 remote desktop sessions. It also adds support for more hardware decoders and switches AV1 decoding to dav1d in supported configurations.

Administrators should identify systems that package or embed FreeRDP, including desktop clients, remote-access gateways, virtual desktop tools, and third-party products built on the library.

Organizations should install FreeRDP 3.31.0 through their supported distribution channel or build the updated release from the official source package. Teams should also verify the downloaded archive using the published SHA-256 checksum and signature where possible.

Prompt patching is especially important for systems that connect to untrusted or internet-exposed RDP servers. The official release includes source archives, ZIP packages, signatures, and checksums for version 3.31.0.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post FreeRDP Fixes 22 Security Flaws and Urges Users to Update Immediately appeared first on Cyber Security News.

Palo Alto Networks has acquired Console, an AI-native platform that will strengthen Cortex by enabling AI-driven security workflows to investigate, prioritize, and remediate threats at machine speed.

The acquisition comes as security teams face rising alert volumes, increasingly automated attacks, and pressure to reduce the time required to detect and contain threats.

Palo Alto Networks said Console’s technology will help organizations use natural-language instructions to create AI agents that can analyze enterprise data, coordinate tasks, and take action across security and IT environments.

The Console is built on the idea that users should be able to define an operational objective. At the same time, AI software handles the technical steps required to complete it.

Palo Alto Networks Acquires Console

In a security operations center, this could include examining an alert, gathering context from endpoint, cloud, identity, and network systems, assessing the severity of the activity, and initiating a response workflow. Palo Alto Networks plans to incorporate Console’s capabilities into Cortex, its security operations portfolio.

The company said the combined platform will help analysts move beyond traditional dashboard-driven operations, where teams manually review alerts and create tickets before taking action.

Nikesh Arora, Chairman and CEO of Palo Alto Networks, said security operations must shift from helping analysts work faster to enabling autonomous security outcomes. He said Console will allow customers to interact directly with enterprise data and create agentic workflows in natural language.

The company described this approach as “software-as-an-agent,” where AI systems can perform operational tasks with the context, permissions, and automation needed to execute security actions.

Rather than functioning only as a copilot that suggests next steps, an agentic system can be configured to complete defined tasks, such as enriching alerts, correlating events, opening or closing cases, isolating endpoints, or triggering remediation processes.

For enterprises, the key challenge will be applying these capabilities safely. Autonomous security agents require controlled access to sensitive systems, detailed logging, human approval for high-impact actions, and safeguards against inaccurate reasoning or unauthorized activity.

Palo Alto Networks said Console will support faster adoption and stronger governance of agentic operations. Andrei Serban, Console’s co-founder and CEO, said the company’s customers have already used agents to reduce operational overhead and improve business processes.

Joining Palo Alto Networks gives Console access to a broader cybersecurity platform, threat intelligence resources, and a global enterprise customer base.

Palo Alto Networks serves more than 70,000 customers and operates across network security, cloud security, security operations, AI, and identity. Its Unit 42 threat intelligence organization will likely provide an important source of context for future AI-driven workflows.

The company cautioned that the benefits of the acquisition depend on successfully integrating Console’s operations and technology into its existing portfolio. It also noted that referenced future services or features may not yet be generally available to customers.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Palo Alto Networks Acquires Console to Build AI Agents for Autonomous Security Operations appeared first on Cyber Security News.

Attackers are actively probing LiteLLM AI gateway deployments for a known authorization flaw that can turn a low-privilege account into full administrative control.

The issue, tracked as CVE-2026-35029, affects LiteLLM versions before 1.83.0 and allows authenticated users to access the sensitive /config/update endpoint without the required administrator role.

LiteLLM acts as an AI gateway between enterprise applications and model providers. It can store provider API keys, database connection details, user data, spending records, and administrative credentials.

This makes exposed LiteLLM control planes a valuable target for attackers seeking cloud credentials, AI service keys, or a path into connected infrastructure. The vulnerability stems from the absence of an authorization check on the /config/update API route.

Hackers Exploit LiteLLM Admin API Flaw

In affected releases, an authenticated account with a limited role, such as the read-only proxy_admin_viewer role, could modify settings reserved for full administrators. LiteLLM corrected the issue in version 1.83.0 by requiring the proxy_admin role for configuration changes.

Researchers observed attackers abusing the flaw to modify environment variables and configuration values. One technique changes UI_LOGO_PATH, a setting that identifies the dashboard logo file.

By replacing the normal image path with a sensitive server-side file, such as /app/.env or /proc/self/environ, an attacker can cause LiteLLM to read it.

The content can then be retrieved through the /get_image endpoint, which reportedly does not require authentication in vulnerable deployments.

This technique can expose secrets stored in environment files and configuration files. Potentially exposed data includes model-provider API keys, the LiteLLM master key, database URLs, AWS credentials, and observability platform tokens.

The flaw can also be used to overwrite UI_USERNAME and UI_PASSWORD environment variables, enabling an attacker to replace dashboard credentials and take over the administration interface.

The security impact can extend beyond data theft. Advisories state that attackers can register malicious pass-through endpoint handlers through altered configuration.

Such handlers may point to attacker-controlled code or infrastructure, creating a route to remote code execution or secret exfiltration.

Zenity honeypot telemetry recorded roughly 3,900 requests against LiteLLM administrative API endpoints between February and June 2026, including about 1,000 requests targeting /config/update.

The first configuration-update probes appeared on April 7, one day after CVE-2026-35029 was publicly disclosed. Researchers later saw direct file-read payloads attempting to load common secret locations, including /app/.env, /home/litellm/.env, /app/config.yaml, and /app/proxy_server_config.yaml.

The activity also included attempts to guess master keys such as sk-1234 and sk-litellm-master-key, generate new API keys, create administrator accounts, enumerate users and keys, probe SCIM provisioning routes, and delete models.

These actions show that threat actors are not merely scanning for the CVE they are mapping and attempting to control exposed AI gateway environments.

Organizations using LiteLLM should immediately upgrade to version 1.83.0 or later. This release adds the missing authorization enforcement for /config/update.nvd.nist+1

Administrators should also remove LiteLLM control-plane services from public internet exposure, place the Admin UI and administrative endpoints behind an authenticated reverse proxy or internal network, and ensure a strong, unique master key is configured. Default or documentation-example values must never be used.

Security teams should review logs for suspicious requests to /config/update, /get_image, /key/generate, /user/new, /model/delete, and /scim/. Requests that set UI_LOGO_PATH to local file paths should be treated as likely attempts at exploitation.

Finally, organizations that operated an exposed LiteLLM version before 1.83.0 should rotate all potentially exposed secrets. This includes LiteLLM master keys, LLM provider API keys, database passwords, cloud credentials, and third-party monitoring tokens, as the vulnerability can expose these values via configuration manipulation and arbitrary file reads.

Learn 7 Metric-Gated AI SOC Deployment Phases – Download Free AI SOC Deployment Playbook 2026.

The post Hackers Exploit LiteLLM Admin API Flaw to Steal Secrets and Target AI Gateway Servers appeared first on Cyber Security News.

SonicWall has warned that attackers are actively exploiting two critical vulnerabilities affecting SMA1000 Series secure mobile access appliances.

The flaws could allow unauthenticated attackers to access sensitive functionality and enable administrators with authenticated access to execute arbitrary operating system commands.

The company published advisory SNWLID-2026-0016 on September 1, 2026, confirming that its Product Security Incident Response Team investigated a case indicating active exploitation.

SonicWall urged organizations to install the available platform hotfixes immediately and review exposed systems for signs of compromise.

The vulnerabilities affect SMA1000 6210, 7210, and 8200v appliances running version 12.4.3-03453 or earlier, as well as version 12.5.0-02835 or earlier. SonicWall stated that SSL-VPN services running on SonicWall firewalls and the SMA 100 Series product line are not affected.

SonicWall RCE Vulnerabilities Exploited

The most severe issue is tracked as CVE-2026-83548 and carries a CVSS score of 10.0. It is a pre-authentication server-side request forgery vulnerability in the SMA1000 Appliance Workplace interface.

According to SonicWall, the flaw stems from an unintended alternate access path that can serve as a forward proxy. A remote, unauthenticated attacker could exploit this path to access sensitive internal functionality and perform unauthorized operations.

The vulnerability is associated with CWE-918, covering server-side request forgery, and CWE-441, which describes an unintended proxy or confused-deputy condition.

SSRF vulnerabilities are especially dangerous in remote-access appliances because they can allow attackers to make requests from the device itself, potentially bypassing network restrictions designed to protect internal services.

SonicWall also addressed CVE-2026-83549, a post-authentication remote code execution flaw in the SMA1000 Appliance Management Console.

The vulnerability has a CVSS score of 7.8 and stems from improper neutralization of special characters in operating system commands.

An authenticated attacker with administrator privileges could exploit the command injection issue to execute arbitrary commands on the appliance operating system.

While this vulnerability requires valid administrator access, it could be especially damaging when chained with another weakness that provides unauthorized access to appliance functions.

Remote-access infrastructure remains a high-value target because it often sits at the edge of enterprise networks and handles user authentication, VPN connectivity, and access to internal resources.

A compromised SMA appliance may provide attackers with a foothold for credential theft, lateral movement, and further network intrusion.

There is no workaround for either issue. Organizations should upgrade SMA1000 appliances to version 12.4.3-03526 or later, or to version 12.5.0-02952 or later, depending on the software branch they have deployed.

SonicWall also recommends contacting technical support to review appliances for indicators of compromise. If compromise indicators are found, organizations should re-image affected physical appliances or redeploy affected virtual appliances.

Administrators should then change all user and administrator passwords and reset TOTP tokens to invalidate potentially stolen authentication factors.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Critical SonicWall Remote Code Execution Vulnerabilities Actively Exploited in Attacks appeared first on Cyber Security News.

HPE has released security updates for HPE Networking Fabric Composer following the discovery of a large set of vulnerabilities that could allow unauthenticated attackers to gain administrator access, run arbitrary commands, and fully compromise affected systems.

The flaws affect HPE Networking Fabric Composer version 7.3.3 and earlier. Fabric Composer is used to manage and automate data-center network fabrics, making a successful compromise particularly serious because the platform can control important network infrastructure.

The most severe vulnerabilities are tracked as CVE-2026-76657 and CVE-2026-76658. Both received a maximum CVSS score of 10.0.

HPE said the API authentication-bypass flaw, CVE-2026-76657, could allow a remote attacker to circumvent existing authentication controls and obtain administrative privileges without valid credentials. This access could lead to a complete takeover of the Fabric Composer host.

CVE-2026-76658 affects the product’s SSH daemon. An unauthenticated remote attacker could exploit the issue to gain administrative access and execute arbitrary commands as a privileged user on the underlying operating system.

HPE Fabric Composer Flaws

In practical terms, a successful exploit may give an attacker control of the appliance and the ability to alter its configuration, steal information, or use it as a foothold for further movement inside an organization’s network.

HPE also fixed CVE-2026-19766, an adjacent-network authentication bypass rated 9.6. The vulnerability could allow an unauthenticated attacker on a connected network segment to execute arbitrary code with privileged operating-system permissions.

Other serious findings include unauthenticated remote code execution bugs, stored cross-site scripting issues, command injection, arbitrary file write, SQL injection, privilege escalation, information disclosure, and denial-of-service flaws.

Several weaknesses are especially concerning because they can be chained. For example, an attacker may first use an information-disclosure bug to understand internal services, then exploit an authentication bypass or remote code execution flaw to take control of the server.

Lower-privileged Fabric Composer users may also be able to exploit API and web interface flaws to escalate to administrative access.

HPE said its internal security researchers discovered the vulnerabilities. At the time the advisory was released, the company said it was not aware of public exploit code or public discussion targeting the issues.

However, the broad range and high severity of the bugs make prompt patching important, particularly for systems whose management interfaces are reachable from untrusted networks.

Organizations using Fabric Composer should upgrade to version 7.4.0 or later in the 7.4 branch, or version 7.3.4 or later in the 7.3 branch.

HPE also recommends restricting command-line and web-based management interfaces to a dedicated Layer 2 segment or VLAN, enforcing Layer 3 firewall controls, and using logging and accounting controls to track access and user activity.

Older releases that have reached End of Maintenance should be treated as potentially exposed unless HPE has explicitly stated otherwise.

Administrators should identify all Fabric Composer installations, confirm their running versions, apply the vendor’s fixes, and review administrator accounts, SSH exposure, API access, and network management logs for suspicious activity.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Critical HPE Fabric Composer Flaws Let Unauthenticated Attackers Execute Code and Take Over Systems appeared first on Cyber Security News.

A critical authentication bypass vulnerability in JFrog Artifactory, tracked as CVE-2026-82329, is being actively exploited, allowing unauthenticated attackers with network access to gain administrator-level privileges.

WatchTowr said its intelligence team has observed attackers exploiting the issue and “minting themselves admin tokens.” An attacker with a valid administrator token could control the affected Artifactory environment, including repositories, user accounts, access permissions, build artifacts, and software packages stored in the platform.

JFrog disclosed the vulnerability on August 28, 2026, and classified it as critical. The company described CVE-2026-82329 as an improper authentication issue, tracked under CWE-287.

Under the default configuration, a remote attacker does not need valid credentials to exploit the weakness and may obtain administrative privileges.

JFrog Artifactory Auth Bypass Exploited

Artifactory is widely used by development and DevOps teams to manage packages, container images, binaries, build dependencies, and other software artifacts.

Because it often sits within CI/CD pipelines, compromising an Artifactory server can pose a serious risk to the software supply chain.

Attackers who gain admin control may be able to alter repository settings, create privileged accounts, steal stored secrets, access private packages, or attempt to introduce malicious artifacts into trusted build and deployment workflows.

According to exposure management firm WatchTowr, the reported creation of administrator tokens is particularly concerning because they can provide persistent access even after an organization changes passwords or terminates active user sessions.

Security teams should investigate whether any unexpected administrator tokens, new privileged users, unusual API activity, or configuration changes were created around the time the vulnerable instance was exposed.

JFrog said its cloud environments have already been fortified, meaning customers using the vendor-managed cloud service do not need to take action for this specific issue.

However, organizations running self-hosted Artifactory must upgrade immediately to a fixed release on their supported branch. The patched versions are 7.111.21, 7.117.28, 7.125.20, 7.133.29, 7.146.38, and 7.161.20.

The affected ranges include Artifactory versions 7.111.4 through 7.111.21, 7.117.0 through 7.117.27, 7.125.0 through 7.125.19, 7.133.0 through 7.133.28, 7.146.0 through 7.146.36, and 7.161.0 through 7.161.19.

Organizations should also restrict external access to Artifactory management interfaces, review reverse-proxy and firewall rules, and ensure only trusted networks can reach administrative endpoints.

Teams should inspect access logs for unfamiliar source IP addresses, failed or abnormal authentication requests, token-generation events, and calls to user, permission, or repository administration APIs.

Administrators should treat any internet-exposed, unpatched self-hosted Artifactory deployment as potentially compromised.

After patching, organizations should revoke and reissue administrator tokens, review all privileged accounts, validate repository integrity, and examine CI/CD credentials that may have been accessible through the platform.

The active exploitation report makes rapid remediation essential. A compromised artifact repository can turn a single authentication bypass into a broader breach affecting developers, build systems, production workloads, and downstream software users.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post JFrog Artifactory Auth Bypass Exploited in Attacks to Gain Admin Access appeared first on Cyber Security News.

WordPress has launched a new security effort that uses artificial intelligence to identify vulnerabilities in its core software before attackers can abuse them.

The initiative comes as the project receives an increasing number of security reports, driven in part by rapid improvements in AI tools for code analysis and vulnerability research. Over the past year, WordPress said incoming vulnerability reports have increased substantially across its ecosystem.

AI models are making it easier for security researchers to inspect large codebases, trace potentially unsafe data flows, and flag risky programming patterns that could otherwise be missed during manual reviews.

The increase in reports is positive for WordPress security, but it also creates a major operational challenge. Every submission must be reviewed, reproduced, validated, prioritized, fixed, tested, and released.

WordPress Using AI to Find Security Flaws

A report may identify a real vulnerability, a configuration-specific issue, a duplicate finding, or a false positive. The WordPress security team must separate urgent flaws from lower-risk issues while ensuring that security patches do not break millions of websites.

The project discussed the issue during its security team meeting at WordCamp US and has now introduced the Core Security Initiative.

The program is built around three priorities: improving the security release process, reducing the backlog of open reports, and using AI-assisted tools to discover vulnerabilities earlier.

The first area focuses on making security releases more predictable. WordPress plans to strengthen automation and end-to-end testing around security updates so fixes can be delivered reliably.

Better testing is especially important for a platform with a large ecosystem of themes, plugins, hosting environments, and custom deployments. A patch must close the vulnerability without causing unexpected failures for site owners.

The second priority is reducing the queue of unresolved reports and known issues. WordPress is adding team members and volunteers to help review findings, confirm impact, and move valid issues through the remediation process. The stated goal is to reduce the number of open security findings to zero.

The third pillar, described as “Crush vulnerabilities with AI,” is the most notable part of the effort. WordPress intends to use AI-assisted scanning and analysis tools to proactively identify security weaknesses rather than waiting for an external researcher or attacker to discover them first.

These tools can help reviewers search for patterns linked to cross-site scripting, privilege escalation, insecure file handling, server-side request forgery, and other common web application risks. However, AI will not replace human security researchers.

Automated tools can identify suspicious code, but expert reviewers are still needed to verify whether a flaw is exploitable, determine its severity, create a safe patch, and test the fix.

WordPress said its AI work will complement, not replace, responsible disclosure reports from the security community. Recent WordPress releases show why faster discovery matters.

Version 7.0.3 addressed multiple security issues, including cross-site scripting, privilege escalation, server-side request forgery, information disclosure, and CSS injection flaws.

Version 7.0.4 later fixed an authenticated remote code execution issue involving malicious file uploads on sites using Imagick and Ghostscript.

WordPress core vulnerability researchers are encouraged to continue reporting findings through the project’s HackerOne program and to follow its reporting guidelines.

For website administrators, the message remains simple: apply WordPress security updates promptly, maintain tested backups, and keep plugins and themes up to date.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post WordPress Is Using AI to Find Security Flaws Before Hackers Can Exploit Them appeared first on Cyber Security News.

A public proof-of-concept exploit has been released for CVE-2026-62911, a Microsoft Exchange Server vulnerability linked to an authentication capture-and-replay weakness.

While Microsoft classifies the issue as an elevation-of-privilege flaw, the published research describes an attack chain that can lead to unauthenticated remote code execution and a SYSTEM-level compromise on vulnerable Exchange environments.

The PoC, published on GitHub by researcher Nguyen Van Hiep, targets the Exchange Mailbox Replication Proxy service, commonly known as MRSProxy.

The exploit documentation claims that a reachable HTTP. sys-hosted MRSProxy endpoint does not enforce Extended Protection for Authentication, enabling an attacker to relay NTLM authentication from an Exchange machine account to the service.

Microsoft disclosed CVE-2026-62911 in August 2026 as an authentication-bypass capture-replay issue affecting on-premises Microsoft Exchange Server.

Public PoC For Exchange Server Pre-Auth RCE

Microsoft’s public description states that an authorized attacker could elevate privileges over a network. In contrast, the available PoC increases practical risk by demonstrating an NTLM relay path that may eliminate the need for prior Exchange credentials in certain configurations.

According to the technical analysis, Exchange exposes MRSProxy through separate endpoints. Extended Protection protects the IIS-hosted /EWS/MRSProxy.svc path.

However, the HTTPsys-hosted MailboxReplicationService ProxyService endpoint may not validate channel bindings. This creates a relay opportunity where a captured machine-account authentication attempt can be forwarded to the vulnerable Exchange service.

The attack chain begins by coercing an Exchange server to authenticate to an attacker-controlled listener through a technique such as PetitPotam. The attacker can then relay the NTLM authentication to a second vulnerable Exchange server.

If the relay succeeds, the machine account is treated as authenticated by MRSProxy. It receives access to Exchange mailbox replication functionality.

The PoC reportedly abuses WCF methods within the replication service, including IMailbox_Config6 and IMailbox_Connect. The first method accepts a file path, while the second may cause Exchange to write content to that location.

An attacker could attempt to place an ASPX webshell into a web-accessible IIS or Exchange virtual directory, then access the file remotely to execute commands under the Exchange service context. Public reporting and technical analysis describe a potential SYSTEM-level outcome.

The vulnerability is associated with a three-bug chain demonstrated by Orange Tsai of DEVCORE at Pwn2Own Berlin 2026. The Exchange compromise earned a $200,000 reward during the competition, underscoring the security impact of chained authentication and service-level weaknesses.

Affected products include Exchange Server 2016 CU23, Exchange Server 2019 CU14 and CU15, and Exchange Server Subscription Edition RTM. Organizations should deploy Microsoft’s August 2026 Exchange security updates immediately.

Fixed builds include Exchange 2016 CU23 (version 15.1.2507.72), Exchange 2019 CU14 (version 15.2.1544.44), Exchange 2019 CU15 (version 15.2.1748.49), and Exchange Server Subscription Edition (version 15.2.2562.46).

Exchange 2016 reached end of support in October 2025, meaning organizations without Extended Security Updates may not have a supported remediation path.

Administrators should also verify Extended Protection settings, reduce unnecessary exposure of Exchange services, monitor for unusual NTLM relay activity, and investigate unexpected ASPX files in Exchange and IIS directories.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Public PoC Released for Microsoft Exchange Server Pre-auth RCE Vulnerability appeared first on Cyber Security News.

Brave has introduced a new Email Aliases feature in desktop browser version 1.94, allowing users to sign up for websites without sharing their real email address.

The privacy-focused feature generates unique forwarding addresses that deliver messages to a user’s primary inbox while keeping that primary address hidden from online services.

The release is the 39th entry in Brave’s ongoing privacy-update series. It was developed by Brave engineers Pavel Beloborodov, Tarik Demirović, Harold Spencer Jr., and DesignOps Lead Agustín Ruiz, with contributions from former Brave privacy engineer Arthur Edelstein.

Email addresses have become valuable identifiers for advertisers, websites, and data brokers. Unlike browser cookies, an email address can follow a person across devices, browsers, and online services.

Companies can use customer email lists to match users with profiles held by advertising platforms, including Google, Meta, and LinkedIn.

Brave Unveils New Email Aliases

For example, a customer may provide an email address while purchasing running shoes from an online store. The store could then upload that address to an advertising platform’s server-side matching system.

If the platform already has the same email linked to a social-media account, it can associate the customer’s purchase with their advertising profile.

This type of data sharing can occur outside the browser, meaning traditional tracker blocking may not stop it. Brave said Email Aliases are designed to reduce this exposure by giving each website a separate address rather than the user’s permanent email identity.

Users can create an alias directly inside an email field on a website. The generated address forwards incoming messages to the email linked with the user’s Brave Account.

If an alias begins receiving spam or is connected to a service the user no longer uses, it can be deactivated and replaced with a new address.

The feature is managed through Settings > Autofill & Passwords > Email Aliases, or through the internal browser page brave://settings/email-aliases.

Users must first create a Brave Account using an email address and password, which is separate from a Brave Premium subscription account.

Brave said its account system uses OPAQUE, a password-authenticated key exchange protocol standardized in RFC 9807. The protocol is intended to ensure that a user’s password is not sent directly to Brave’s servers during authentication.

According to Brave, it stores the primary account address and generated aliases encrypted at rest. The company said it does not read email content, but processes incoming mail for spam and malware filtering before forwarding it. Messages are deleted from Brave’s servers within seconds after delivery.

Alias notes are stored locally on the user’s device. When Brave Sync is enabled, those notes are encrypted end-to-end between devices in the same Sync chain.

Brave is initially offering five free email aliases per user. The company said it plans to expand the capability to mobile devices and introduce a Premium version later. It also warned that some forwarded messages may initially reach spam folders while the service establishes its email-sending reputation.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Brave Unveils New Email Aliases to Keep Your Personal Email Address Private appeared first on Cyber Security News.

Google has introduced new privacy and network-security protections in Android 17, including stricter controls that prevent apps from scanning devices connected to a user’s local Wi‑Fi network without permission.

The update is designed to reduce household profiling, limit tracking, and defend users against threats ranging from malicious Wi‑Fi activity to cellular SMS scams.

Android smartphones often connect to home Wi‑Fi networks shared by smart TVs, cameras, game consoles, speakers, printers, and other Internet of Things devices.

Previously, an app could potentially discover devices on the same local network without clearly asking the user for access. That information could reveal details about a household’s technology, behavior, or device ownership.

Android 17 now enforces Local Network Protection, requiring apps to request explicit permission before scanning for or connecting to devices on the local network.

This gives users more control over which applications can view nearby devices and communicate across their home Wi‑Fi environment. The restriction does not remove common features such as streaming content to a television.

Instead, Google recommends that developers use secure Android system tools for actions such as casting. Users can select a compatible TV or device through the operating system, without giving the app broad visibility into every device connected to the network.

Android 17 New Protection for Wi-Fi Tracking

The Wi‑Fi privacy improvement is part of a wider Android 17 security update focused on protecting network metadata and encrypted connections.

Google is also adding support for Encrypted Client Hello, or ECH, a privacy technology that hides the name of a website a device is trying to reach during the early stages of an HTTPS connection.

HTTPS already encrypts most traffic between a device and a website. However, network operators and eavesdroppers can sometimes still see domain-name information, allowing them to identify websites or apps being accessed.

This metadata can be used for profiling or to support targeted phishing and scam operations. With ECH and Private DNS, Android 17 aims to encrypt more of this connection data.

For supported websites and applications, the change makes it more difficult for Wi‑Fi operators, internet providers, and network snoops to determine a user’s online destination.

Google’s Android 17 enables Certificate Transparency by default, requiring website and app certificates to be recorded in public logs to help detect suspicious or improperly issued certificates that could enable interception attacks.

Another major change targets 2G downgrade attacks and SMS blaster activity. Criminals use fake cellular base stations to force nearby devices onto older 2G networks, where security protections are weaker.

Attackers can then send fraudulent text messages that imitate banks, delivery companies, government agencies, or other trusted organizations. Android previously offered users a manual option to turn off 2G connectivity.

Android 17 expands this defense by allowing participating mobile carriers to turn off 2G by default for subscribers. The zero-click protection reduces exposure to rogue base stations and phishing messages delivered through legacy cellular networks.

For users, the changes mostly operate in the background. For developers, Android 17 introduces new requirements for local network access and modern encrypted networking practices. Together, the measures show Google’s effort to reduce data exposure across Wi‑Fi, web browsing, certificate validation, and cellular communications.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post Android 17 Adds New Protection for Wi-Fi Tracking to Keep Your Device Secure appeared first on Cyber Security News.

OpenClaw has released version 2026.8.1, also called OpenClaw 2.0, in what the open-source AI agent platform described as its largest update to date.

The release was built by 933 contributors, including 569 first-time contributors, and contains more than 16,000 pull requests across installation, agents, plugins, credentials, browser controls, messaging, automation, memory, and native applications.

The nearly two-month development cycle marks a major shift for the project, which previously shipped 106 releases in 230 days.

OpenClaw said it paused its usual rapid release cadence because the growing project needed a stronger technical foundation and safer upgrade path for both new and existing deployments.

A key focus of OpenClaw 2.0 is security for AI agents that can access tools, files, browser sessions, messaging platforms, cloud workers, and enterprise services.

The update introduces private credential requests, allowing an agent to request a secret via a masked prompt without exposing the credential value in the chat history or model context.

An opt-in proxy can also restrict protected-secret substitution to approved destinations, helping reduce the risk of credentials being exposed through unintended outbound requests. The platform now includes a shared credential store for team environments.

OpenClaw 2.0 Released

Administrators can manage team-scoped secrets and environment values through SQLite-backed CLI and Settings interfaces. Secret values remain write-only, while protected outbound connections can be bound to declared hosts.

OpenClaw also added an optional 1Password broker that supports curated secret references, service-account authentication, per-secret approval, and audit records without exposing the secret value.

Plugin security also received major upgrades. OpenClaw now presents capability, source, version, and artifact details before external plugins are installed or enabled. Installations from arbitrary executable sources require the –force flag.

In contrast, trusted ClawHub, bundled, official-catalog, and tracked-update sources can avoid the provenance warning but still require capability consent. The release also adds ClawHub security-audit information to the plugin installation flow.

For agent execution, OpenClaw 2.0 introduces explicit session permission modes and workspace restrictions. Restricted filesystem access is anchored to the recorded workspace or worktree, reducing the chance that an agent can access files outside its approved scope.

Team operator roles can limit which agents, sessions, and administrative scopes are available to verified users. However, OpenClaw warns that these controls are collaboration features and should not be treated as hostile multi-tenant isolation.

The release also improves approval handling for recurring automations. Users can approve a specific operation once, inspect or revoke the permission later, and require a new approval when the automation’s operation changes. This helps prevent an initially approved workflow from silently expanding its authority over time.

Other defensive improvements include model allowlists, configuration-change history with sensitive-value redaction, database recovery protections, sanitized debugging handoffs through OpenClaw triage, safer startup migrations, and fixes designed to prevent private prompt context from appearing in final or streaming replies.

OpenClaw 2.0 also rebuilds the browser Control UI as a primary workspace and adds shared cloud sessions, browser workflow controls, agent dashboards, and broader support for local and external model providers.

The scale of the update makes security review and staged deployment especially important for organizations using OpenClaw agents with production credentials, plugins, messaging integrations, or cloud execution.

Prevent incidents due to slow investigations. Power your Tier 1 with threat intelligence from 15K SOCs: Integrate TI Lookup in your SOC

The post OpenClaw 2.0 Released With Major Security Upgrades for AI Agents, Plugins and Credentials appeared first on Cyber Security News.