The feature is designed to address the growing use of more convincing scams, including messages created with the help of AI. Instead of sending messages to WhatsApp’s servers for analysis, Scam Alert uses a small machine learning model that runs directly on the user’s device.

Once enabled, the model examines messages from people who are not saved as contacts. It looks for patterns in the conversation and language that may indicate common scam techniques.

The message itself stays on the device during this process. WhatsApp says it does not automatically send messages to Meta, WhatsApp, or another third party for review. Users decide what happens next.

If a message appears suspicious, WhatsApp can display a warning to the recipient. The user can then choose to block the sender, report the conversation, continue chatting, or mark the conversation as trusted if they believe the warning is incorrect.

Privacy Is a Key Part of Scam Alert

WhatsApp says the system was designed around three main ideas: local processing, no automatic reporting, and user control.

The company still needs some information to understand how well the feature performs. Instead of collecting individual messages, WhatsApp uses a privacy-focused analytics system that gathers limited information such as how many warnings were displayed and what actions users took.

This information is processed using Trusted Execution Environments (TEEs) and additional privacy techniques before aggregated statistics are sent to Meta.

Protecting the AI Model

Another concern is making sure attackers cannot secretly deliver a modified version of the scam-detection model to specific users.

WhatsApp says each model version is published with a SHA-256 hash in an append-only transparency system before it is deployed. This creates a record that can be used to verify that the model has not been secretly changed.

Model downloads also use an Oblivious HTTP (OHTTP) relay, which helps prevent the server from directly linking a model request to a user’s IP address.

WhatsApp says even the process used to assign users to different model versions for testing happens locally on their devices rather than being controlled by the server.

Additional Security Controls

The company says Scam Alert was designed to protect against several types of threats, including outside attackers, malicious employees, and compromised third-party suppliers.

Its security measures include isolated confidential computing environments, encrypted memory, and additional protections around the systems running the analytics infrastructure.

Users can also check information about the feature through WhatsApp’s transparency controls. The in-app activity section shows details such as which messages were analyzed and which model version was used.

External Researchers Can Test the System

WhatsApp is also expanding its Bug Bounty program to cover parts of the Scam Alert technology, including the machine learning models and analytics infrastructure.

This gives security researchers an opportunity to look for weaknesses and verify whether the system is being used only for its stated purpose of detecting scams.

Limited Beta Release

Scam Alert is initially being introduced through a limited beta rollout. WhatsApp says it plans to continue testing the technology with security researchers before making it more widely available.

The company also plans to publish a technical white paper explaining how the system works in greater detail.

The approach reflects a growing focus on building AI-powered security features without giving up user privacy. Instead of sending private conversations to the cloud for analysis, WhatsApp is attempting to combine on-device AI, confidential computing, and transparency mechanisms to detect scams while keeping message content protected.

The post WhatsApp Scam Alert: New Protection Against Hackers appeared first on First Hackers News.

The research involves OpenAI, Anthropic, and Google, and shows that encrypted reasoning data produced by powerful models could potentially be reused with less restricted models from the same provider. In some cases, this could allow the hidden reasoning to be reconstructed in readable text.

Researchers from the ELLIS Institute Tübingen, Max Planck Institute, MATS Research, and Snyk investigated the issue across the Claude, GPT, and Gemini ecosystems. The attack does not require special privileges and can be carried out through normal API access.

How the Attack Works

Modern reasoning models perform additional processing before producing their final answers. Much of this internal reasoning is intentionally hidden from users because it may contain sensitive information, system instructions, safety-related decisions, or proprietary model behavior.

Instead of keeping all conversation state on the server, some AI APIs return encrypted data that can be sent back with later requests. This allows the model to continue a conversation without storing every detail on the provider’s side.

Researchers found a potential weakness in how these encrypted reasoning packages were protected.

The data was not sufficiently tied to the specific model, user, or session that originally created it. This meant a reasoning package generated by a more heavily protected model could potentially be submitted to another model within the same provider’s ecosystem.

Using Smaller Models to Reveal Hidden Data

The researchers demonstrated an attack technique in which a reasoning package from a powerful model was passed to a less restricted model.

The smaller model could then be prompted to reproduce information contained inside the encrypted package.

This creates an unusual security problem. The attacker does not necessarily need to break the encryption directly. Instead, another model can potentially act as an intermediary that helps expose the information.

The research team reported similar behavior across the GPT, Claude, and Gemini model families.

Researchers also compared the recovered reasoning data with usage information provided through the APIs to determine whether the extracted material corresponded to the original reasoning process.

Hidden Reasoning Can Contain Sensitive Information

The concern goes beyond exposing how an AI model thinks.

Researchers analyzed thousands of publicly available AI agent transcripts collected from sources such as GitHub and Hugging Face. Their analysis identified hundreds of pieces of potentially sensitive information inside recovered reasoning data.

This included:

• 367 pieces of personally identifiable information
• 182 hardcoded credentials
• 62 API keys
• 33 passwords
• 30 personal email addresses

The problem is that some of this information may never appear in the model’s visible response.

A developer could therefore publish an AI session or agent log believing it contains no secrets, while sensitive information remains hidden inside associated reasoning data.

A New Risk for AI Agents

The research also highlights a potential problem for autonomous AI systems.

If security monitoring only examines the visible conversation between a user and an AI agent, malicious instructions hidden inside reasoning-related data could potentially escape detection.

An attacker could attempt to place instructions inside an encrypted reasoning package and have those instructions processed later by an AI agent.

This creates a potential indirect prompt injection scenario where the malicious content is not obvious in the conversation that security tools are monitoring.

For organizations using AI agents to interact with cloud services, databases, code repositories, or business applications, this type of hidden input deserves particular attention.

Which AI Platforms Were Studied?

The research covered major model ecosystems from three providers:

The exact attack paths and model combinations varied between providers, but the broader issue was similar: reasoning data was not sufficiently isolated from other models within the same ecosystem.

Providers Have Already Responded

The researchers reported the findings to the affected companies through responsible disclosure.

OpenAI, Anthropic, and Google acknowledged the research and introduced server-side protections. According to the researchers, the original proof-of-concept attacks could no longer be reproduced against the updated API implementations.

That reduces the immediate risk from the specific attack demonstrated in the research, but the findings highlight a larger issue for the AI industry.

As AI systems become more complex, security controls need to protect not only visible prompts and responses but also the internal data exchanged between models and supporting services.

How Organizations Can Reduce the Risk

AI providers and companies building applications around LLMs can take several steps to protect sensitive reasoning-related data.

Bind encrypted data to its source. Reasoning packages should be cryptographically linked to the specific model, user, and session that created them.

Separate model tiers. A payload generated by one model should not automatically be accepted by another model unless that behavior is explicitly intended and securely controlled.

Rotate older cryptographic keys. Where historical encrypted data may have been exposed, organizations should consider key rotation and invalidation strategies.

Protect AI logs. Developers should treat reasoning-related payloads, signatures, and encrypted model data as sensitive information. These fields should be removed or sanitized before logs are shared publicly.

Why This Matters

The research shows that protecting AI systems is not only about securing the model itself.

Modern LLM platforms involve APIs, model tiers, encrypted payloads, agent frameworks, logging systems, and multiple supporting services. A weakness in the connection between these components can create a security problem even when the underlying model remains protected.

For organizations adopting AI agents and reasoning models, the lesson is clear: hidden data should be treated as sensitive data, even when users cannot see it.

The post LLM API Bugs Leak Hidden Reasoning Traces appeared first on First Hackers News.

The update was released on August 11, 2026, and security teams have another important reason to move quickly: three of the vulnerabilities were already publicly known or are being actively exploited, according to Microsoft’s advisory information.

For organizations with large Microsoft environments, this is a release that should be reviewed early rather than left for the normal monthly patching cycle.

Three Vulnerabilities Need Immediate Attention

The most important issues this month involve Windows components.

CVE-2026-72971 affects the Windows Container Isolation File System Filter Driver (unionfs.sys). Microsoft rates it as Important and says the vulnerability was publicly disclosed before the security update was released. The issue could allow an attacker to tamper with protected system or container-related data.

Another publicly disclosed issue, CVE-2026-62832, affects the Windows User Profile Service. It is an elevation-of-privilege vulnerability, meaning an attacker with an existing foothold could potentially use it to obtain additional permissions on a vulnerable machine.

The biggest concern is CVE-2026-68820. Microsoft lists this Windows WinSock-related vulnerability as actively exploited in the wild. Because attackers are already using the flaw, organizations should give this update priority and investigate potentially affected endpoints.

Where the 394 Vulnerabilities Fall

The vulnerabilities fixed this month cover several major security categories:

The numbers show that privilege escalation and remote code execution continue to make up a significant portion of Microsoft’s monthly security workload.

For defenders, this means patching should not focus only on vulnerabilities carrying a Critical rating. An Important vulnerability can still become dangerous when it is combined with another weakness or used after an attacker gains initial access.

SharePoint Remains a Key Target

Several SharePoint Server vulnerabilities are included in the August release.

Among them are CVE-2026-70355, CVE-2026-70326, and CVE-2026-70324, which involve privilege escalation, as well as CVE-2026-70321, a remote code execution vulnerability.

SharePoint environments deserve special attention because they often sit close to valuable corporate information. A compromised server could potentially give attackers access to documents, accounts, internal applications, or other parts of an organization’s environment.

Companies running internet-accessible SharePoint servers should therefore make these systems a priority during their patching process.

Azure and Windows Infrastructure Also Affected

The update includes CVE-2026-71331, a Critical remote code execution vulnerability involving Microsoft Azure Attestation and Device Health Attestation services.

Issues affecting cloud and device-trust infrastructure can have broader implications because these services may be part of an organization’s authentication and security architecture.

Windows infrastructure is also covered by several fixes, including vulnerabilities affecting DNS, Windows Management Services, Windows Installer, and other system components.

Organizations should identify which of these services are actually deployed in their environments and prioritize systems that are externally accessible or support critical business operations.

Don’t Forget PowerShell and Developer Machines

Microsoft has also patched vulnerabilities in PowerShell Core and .NET.

One PowerShell issue, CVE-2026-70337, could lead to remote code execution, while CVE-2026-70338 involves a security feature bypass.

This is particularly relevant to security teams because PowerShell is both an essential administration tool and a common technique used during attacks. Keeping the platform patched should go hand in hand with monitoring suspicious PowerShell activity.

Developer environments should also be included in patching plans. Vulnerabilities affecting Visual Studio Code and GitHub Copilot could create additional risk when compromised developer machines have access to source repositories, cloud credentials, or CI/CD infrastructure.

A developer endpoint is not just another workstation. In many organizations, it can provide a pathway into systems that are critical to software development and production deployment.

Microsoft Office Is Part of the Release

Office applications receive their share of security fixes as well.

The August updates cover Outlook, Word, Excel, PowerPoint, and Microsoft Office, including remote code execution and information-disclosure vulnerabilities.

These issues deserve attention because attackers frequently use Office documents and email as an entry point for phishing and malware campaigns. A vulnerability that appears less urgent in isolation can become much more serious when combined with social engineering.

What Security Teams Should Do Now

The size of this month’s release can make patching 394 vulnerabilities seem overwhelming. The practical approach is to prioritize based on exposure and exploitability.

Start with the vulnerabilities that are actively exploited or publicly disclosed. Next, focus on Critical vulnerabilities and systems exposed to the internet, particularly SharePoint and important Azure or Windows infrastructure.

After that, review developer workstations, PowerShell installations, Office applications, and other Microsoft products used across the organization.

Teams should also use this update as an opportunity to check whether vulnerable systems are being properly identified through asset inventories and vulnerability-management tools.

The key takeaway from Microsoft’s August 2026 release is simple: don’t treat Patch Tuesday as just another routine update cycle. With 394 vulnerabilities addressed and an actively exploited Windows flaw among them, organizations should move quickly to identify their exposure and deploy the appropriate security fixes.

The post 394 Microsoft Flaws Patched in August Update appeared first on First Hackers News.

The most critical vulnerability, tracked as CVE-2026-53413, has been named “Zoomsday” by A Security, the research team that discovered the flaw. Zoom has rated the vulnerability as high severity.

The problem is linked to Zoom’s annotation feature, which allows participants to draw, highlight, and add text while someone is sharing their screen. Behind the scenes, this feature uses a proprietary communication protocol to exchange data between participants.

Researchers found that Zoom’s annotation component does not properly validate certain formatting information received over the network. Specifically, the function responsible for processing annotation data uses fixed-size memory buffers while trusting size values provided by the sender.

An attacker could take advantage of this weakness by sending specially crafted data that is larger than the allocated memory space. This can cause a buffer overflow, potentially allowing the attacker to corrupt memory and take control of the application.

A Security demonstrated the issue on macOS by using the vulnerability to launch Safari on a targeted computer without the user taking any action. The demonstration showed that successful exploitation could occur quietly, without obvious signs that the device had been compromised.

The attacker only needs to join or host a Zoom meeting and target another participant.

Three More Zoom Vulnerabilities

Zoom also disclosed three additional security issues as part of the same security update cycle.

CVE-2026-53414 is a medium-severity buffer over-read vulnerability affecting Zoom Clients. Successful exploitation could allow an attacker to access information from areas of memory that should not be exposed.

CVE-2026-53415 is a high-severity use-after-free vulnerability. It can cause memory corruption and could potentially lead to code execution when an attacker causes the application to access memory that has already been released.

CVE-2026-53416 affects the Zoom VDI Client and is caused by a path traversal vulnerability. An attacker could manipulate file paths to access files outside the directories intended by the application.

Zoom published these issues under security bulletins ZSB-26015 through ZSB-26018, with the bulletins published and updated on August 11, 2026.

Zoom Releases Security Fixes

Zoom says the annotation-related vulnerabilities affect its clients across supported platforms, while the path traversal issue is limited to VDI environments.

The company has already released fixes. The affected products should be updated to the following versions:

• Zoom Workplace: 7.1.5 or 7.0.6
• Zoom Rooms: 7.1.5
• Meeting SDK: 7.1.5
• Workplace VDI Client: 7.0.11 or 6.6.16
• VDI Plugin: 7.0.11 or 6.6.15

What Organizations Should Do

Zoom has not reported any confirmed exploitation of these vulnerabilities in the wild, and there is currently no publicly available proof-of-concept exploit.

However, the zero-click nature of CVE-2026-53413 makes the issue particularly important. Successful exploitation does not require the victim to open a file, click a link, or approve a prompt.

Organizations should prioritize updating Zoom across managed devices as soon as possible. IT and security teams should also consider deploying the updated packages centrally rather than depending entirely on users to update their applications themselves.

Keeping centralized installations updated can help prevent vulnerable versions from being reintroduced during future software deployments.

The post Zoom Zero-Click Flaws Put Devices at Risk appeared first on First Hackers News.

According to recent security research, while the overall number of blocked Android malware attacks declined during the second quarter of 2026, banking malware remains one of the most active mobile threats. Rather than disappearing, attackers are changing how their malware is packaged and deployed.

A Shift Toward Dropper-Based Attacks

A dropper is an application that appears harmless but is designed to download or activate malware after installation.

This approach allows attackers to hide the real banking Trojan during the initial app review process. Once the application is installed on a victim’s device, it can retrieve additional malicious components and begin targeting banking credentials.

By separating the delivery mechanism from the actual malware, cybercriminals can update or replace their payloads without creating entirely new malicious applications.

Malicious Apps Disguised as Legitimate Software

Researchers observed attackers disguising malware as legitimate Android applications, including utility and document reader apps.

In one campaign, a PDF reader displayed what appeared to be a routine software update notification. Instead of installing an update, the application downloaded banking malware onto the victim’s device.

These fake update prompts make malicious activity appear normal, increasing the likelihood that users will unknowingly install the malware.

Smarter Delivery Techniques

Modern Android droppers are becoming more selective in how they deliver malware.

Some applications first collect information about where they were downloaded and send that data to a command-and-control (C2) server. The server decides whether to deliver the malicious payload based on the installation source.

This selective delivery helps attackers avoid security researchers, automated analysis tools, and app store review systems while targeting real users.

Banking Malware Continues to Evolve

Security researchers also observed continued activity from well-known Android banking malware families, including Mamont and Creduz.

New variants are being released regularly, suggesting that malware operators are continuously testing new delivery techniques, improving evasion methods, and developing updated versions to avoid detection.

Why This Matters

The growing use of droppers shows that mobile threats are becoming more sophisticated. A reduction in traditional banking Trojan detections does not necessarily indicate a lower risk—it may simply reflect changes in how malware is delivered.

As attackers continue refining their techniques, users and organizations should remain cautious of applications that request unexpected updates or unnecessary permissions.

How to Stay Protected

To reduce the risk of Android banking malware:

• Install apps only from trusted sources.
• Keep Google Play Protect enabled.
• Avoid downloading apps from unofficial websites.
• Be cautious of unexpected in-app update requests.
• Review requested permissions before installing applications.
• Keep Android devices and applications updated.
• Use a reputable mobile security solution.

Conclusion

Android banking malware operators are shifting away from traditional delivery methods and increasingly relying on dropper applications to bypass security controls. As these techniques become more advanced, mobile users and organizations should strengthen their security practices, verify application sources, and remain alert to suspicious app behavior to reduce the risk of financial compromise.

The post Android Banking Malware Adopts New Evasion Techniques appeared first on First Hackers News.

Rather than breaking Windows Hello encryption, the technique takes advantage of an already authenticated Windows session. If an attacker gains access to a logged-in device, they may be able to leverage the victim’s existing Windows Hello credentials to request authentication tokens and potentially establish long-term access to Microsoft Entra resources.

The research highlights how attackers can abuse trusted authentication mechanisms once an endpoint has already been compromised.

How the Attack Works

Windows Hello for Business replaces passwords with cryptographic keys that are securely stored on the user’s device, typically inside the Trusted Platform Module (TPM). These keys are designed to remain protected and cannot normally be exported from the device.

However, the research demonstrates that an application running with standard user privileges inside an active Windows session can request cryptographic operations from the Windows Hello key without asking the user to re-enter their PIN or biometric authentication.

Instead of stealing the private key itself, the attacker simply instructs Windows to use it on their behalf. As long as the user remains logged in, Windows considers the authentication request valid.

This effectively allows attackers to “borrow” the trusted Windows Hello credentials without directly compromising them.

Authentication Without Traditional Credentials

Researchers showed that the borrowed Windows Hello key can be used during Microsoft’s WebAuthn authentication process.

Since the authentication challenge is not permanently tied to a specific device or session, attackers can generate the request from their own system while having the victim’s compromised device perform the required cryptographic signing.

‍Follow Us on:Linkedin, Instagram, Facebook to get the latest security news!

Once the challenge is successfully signed, Microsoft Entra treats the authentication as legitimate and issues cloud authentication tokens.

Unlike earlier attack techniques, this approach removes the need for attackers to control another Microsoft Entra-registered device, making exploitation significantly easier.

Potential Impact

Successful exploitation could provide attackers with access to Microsoft Entra cloud resources while using legitimate authentication mechanisms.

Depending on the organization’s security configuration, attackers may be able to:

• Authenticate to Microsoft Entra services.
• Obtain cloud authentication tokens.
• Register attacker-controlled devices.
• Add new authentication methods.
• Maintain persistent access to cloud identities.
• Expand access across enterprise environments.

Because Windows Hello for Business satisfies multi-factor authentication requirements, attackers may also bypass additional protections designed to secure identity management functions.

Why Organizations Should Pay Attention

The research demonstrates that passwordless authentication is only as secure as the endpoint itself.

Even though Windows Hello protects credentials from theft, an attacker who compromises an active Windows session may still abuse trusted authentication processes without ever knowing the user’s password or PIN.

This reinforces the importance of endpoint security alongside identity protection.

Organizations should not rely solely on passwordless authentication but should also focus on preventing attackers from gaining access to active user sessions.

Detection and Mitigation

Security teams should monitor Microsoft Entra sign-in activity for unusual Windows Hello authentication events, particularly those that do not contain an associated device ID.

Administrators should also investigate unexpected device registrations, monitor changes to authentication methods, strengthen endpoint detection and response capabilities, and regularly review Conditional Access policies to ensure only trusted devices can register new authentication factors.

Monitoring active user sessions and detecting suspicious endpoint activity remain critical for preventing attackers from abusing trusted authentication mechanisms.

Conclusion

Mollema’s research demonstrates that compromising an active Windows session can have serious consequences, even in environments that have adopted passwordless authentication.

While the technique does not break Windows Hello’s cryptographic protections, it shows how trusted authentication workflows can be abused after an endpoint has been compromised.

Organizations using Windows Hello for Business and Microsoft Entra should prioritize endpoint protection, continuous monitoring, and identity security controls to reduce the risk of attackers turning temporary access into persistent cloud compromise.

The post Windows Hello Key Flaw Opens Door to Entra ID Access appeared first on First Hackers News.

The attackers are primarily targeting users through Discord servers and gaming forums, where fake cheat downloads are shared as legitimate software. By taking advantage of the popularity of Roblox cheats, the campaign aims to steal sensitive information and gain complete control of victims’ computers.

Researchers say the operation continues to evolve, with new infrastructure and malware capabilities being added regularly.

How the Attack Works

The infection starts when users download what appears to be a genuine Xeno Roblox cheat.

The downloaded archive looks convincing, containing realistic folder structures and harmless-looking files that make it appear authentic.

Instead of launching a game cheat, the installer quietly begins executing malicious code in the background.

If Java is not already installed on the system, the malware automatically installs a local Java Runtime Environment without the user’s knowledge. This prepares the system for the next stage of the attack.

Multi-Stage Malware Deployment

After the initial infection, the malware loads a heavily obfuscated Java application disguised as a normal Windows executable.

Before continuing, it performs several checks to determine whether it is running inside a virtual machine, sandbox, or debugging environment. These techniques help attackers avoid detection by security researchers.

‍Follow Us on:Linkedin, Instagram, Facebook to get the latest security news!

The malware then collects basic system information and securely communicates with its command-and-control (C2) server to register the infected device and download additional malicious components.

Java RAT Gains Full System Access

The final payload is a Java Remote Access Trojan (RAT) hidden inside folders designed to resemble legitimate Microsoft GameDVR files associated with the Xbox Game Bar.

To remain active after a reboot, the malware creates registry Run entries using names that appear to be legitimate Windows components. It also attempts to obtain elevated privileges, allowing it to perform more advanced malicious activities.

Once established, the RAT connects to attacker-controlled servers and waits for further instructions.

What Information Does the Malware Target?

Unlike basic information-stealing malware, this campaign combines credential theft with powerful remote surveillance features.

The malware is capable of:

• Stealing saved passwords and browser cookies.
• Collecting credentials from Chrome, Edge, Opera, and Brave.
• Hijacking Discord, Roblox, and Minecraft accounts.
• Targeting cryptocurrency wallets, including Exodus.
• Capturing keystrokes and mouse activity.
• Taking screenshots.
• Streaming the victim’s desktop.
• Accessing webcam feeds.
• Uploading, downloading, and modifying files.
• Running PowerShell commands.
• Providing attackers with remote shell access.

These capabilities allow attackers to fully control an infected computer while collecting valuable personal and financial information.

Why Gamers Are Being Targeted

Researchers believe the campaign specifically targets Roblox players because many users search online for “free” or “undetected” cheat tools.

Young gamers are especially at risk, as they may download unofficial software from Discord communities or third-party websites without realizing it contains malware.

Since many gaming PCs are shared with family members, a successful infection could also expose banking information, personal documents, saved passwords, and private communications stored on the same device.

Malware Campaign Continues to Evolve

Security researchers, including Bitdefender and previous investigations by ThreatLocker, have linked the campaign to malware previously tracked as Powercat.

The operation has reportedly been active since early 2026 and continues to expand through new command-and-control servers and updated malware modules. The attackers also use encrypted communications and in-memory payload execution, making the malware more difficult to detect and remove.

How to Stay Protected

To reduce the risk of infection, users should follow these security best practices:

• Avoid downloading unofficial Roblox cheats or game modification tools.
• Only install software from trusted sources.
• Keep antivirus and security software up to date.
• Enable multi-factor authentication (MFA) on gaming and email accounts.
• Regularly update Windows and installed applications.
• Be cautious of download links shared through Discord servers or online gaming forums.

As cybercriminals increasingly target gaming communities, staying away from unofficial cheat software remains one of the most effective ways to avoid malware infections and protect personal information.

IOCs

The post Fake Roblox Hacks Target Discord and Gaming Credentials appeared first on First Hackers News.

According to security researchers at AISLE, the flaw could have allowed attackers to compromise a developer’s system simply by convincing them to click a malicious link embedded in a Git commit message.

The issue has now been fixed across all affected platforms, but it highlights the growing security risks facing modern AI-powered development tools.

How the Attack Worked

The vulnerability relied on a malicious link hidden inside a Git commit message.

If a developer clicked the link while viewing the commit inside one of the affected editors, arbitrary code could run automatically with the same permissions as the developer.

The attack required only a single click and did not display any warning, confirmation message, or security prompt, making the compromise difficult to notice.

What Attackers Could Do

Successful exploitation could have given attackers extensive control over a developer’s machine.

Potential impacts included:

• Stealing API keys and other sensitive credentials.
• Installing malware or keyloggers.
• Accessing, modifying, or deleting local files.
• Maintaining persistent access even after the editor was closed.

Because the malicious activity could remain hidden, developers might not realize their systems had been compromised.

Vulnerability Found Across Multiple Editors

AISLE first discovered the flaw in Microsoft VS Code during security testing in late 2025.

Since Cursor is built on the VS Code codebase, it inherited the same vulnerability. Later, researchers identified the identical issue in Google Antigravity, another AI-assisted coding environment based on the same architecture.

The discovery shows how a single vulnerability in a shared codebase can spread across multiple developer tools used by millions of people.

Vendors Released Security Fixes

After receiving responsible disclosure reports, all three vendors addressed the issue.

• Cursor released a security update shortly after notification.
• Google patched the vulnerability in Antigravity within days.
• Microsoft later released a fix for VS Code.

Current versions of all three editors are no longer affected.

Why This Matters

Many modern AI coding tools are developed using common open-source foundations such as VS Code. While this speeds up development and feature delivery, it also means a single security flaw can impact multiple products simultaneously.

This incident demonstrates the importance of continuous security testing and timely patch management, particularly for AI-powered development environments.

What Developers Should Do

Developers and organizations should take the following steps:

• Update VS Code, Cursor, or Google Antigravity to the latest version.
• Review recent Git activity for anything suspicious.
• Rotate API keys and credentials if older vulnerable versions were used.
• Monitor developer systems for unusual behavior or unauthorized access.

Keeping development tools fully updated and regularly reviewing credentials can help reduce the risk of future software supply chain attacks.

The post Critical Vulnerability Disclosed in VS Code, Cursor, and Google Antigravity appeared first on First Hackers News.

The vulnerability has been assigned a CVSS v4.0 score of 8.7, highlighting the importance of applying the latest firmware updates as soon as possible.

Buffer Overflow Vulnerability Identified

The issue is caused by a stack-based buffer overflow in the router’s Real-Time Streaming Protocol (RTSP) connection-tracking component.

The vulnerability can be triggered when a device connected to the local network communicates with a malicious RTSP server. If the server responds with specially crafted data, it can cause improper memory handling inside the router, leading to a system crash or remote code execution.

Unlike many router vulnerabilities, this flaw does not require attackers to authenticate with the device under its default configuration.

However, exploitation requires some user interaction. A user must first access a malicious RTSP stream, such as opening a harmful media link or connecting to an attacker-controlled streaming service.

Potential Security Risks

If successfully exploited, attackers could:

• Execute malicious code on the router.
• Modify network or security settings.
• Monitor or intercept network traffic.
• Install persistent malware.
• Use the compromised router to launch attacks against other devices on the network.

Because routers act as the gateway between users and the internet, compromising one can expose every connected device to additional risks.

Affected Versions and Fixes

The vulnerability only impacts TP-Link TL-WR940N hardware version v6.

TP-Link has released firmware updates that address the issue for regional versions, including:

• (EN)_V6_260528
• (US)_V6_260528
• (JP)_V6_260527

Users should download and install the appropriate firmware from TP-Link’s official support website as soon as possible.

Recommended Security Measures

Until systems are updated, users and organizations should:

• Install the latest TP-Link firmware.
• Avoid connecting to untrusted RTSP streaming services.
• Secure the router’s administration interface with a strong, unique password.
• Monitor network activity for unusual behavior.
• Regularly review router firmware for new security updates.

Conclusion

The discovery of CVE-2026-12935 highlights the importance of keeping networking devices up to date. Since routers are a critical part of any home or enterprise network, timely firmware updates and good security practices remain essential to protecting against remote attacks and maintaining a secure network environment.

‍Follow Us on:Linkedin, Instagram, Facebook to get the latest security news!

The post Critical TP-Link Router Flaw Enables Remote Code Execution appeared first on First Hackers News.

To help reduce the risk of attacks before users install the update, Google is temporarily limiting technical details about many of the vulnerabilities until the rollout is complete.

370 Security Flaws Addressed

This release includes fixes for vulnerabilities affecting several core Chrome components, including graphics, networking, browser rendering, and platform-specific features.

Many of these issues were discovered by Google’s internal security teams using automated testing tools designed to detect memory errors and other software weaknesses before they could be exploited.

Seven Critical Vulnerabilities Fixed

Chrome 151 resolves seven critical vulnerabilities, tracked as CVE-2026-17650 through CVE-2026-17656.

Several of these are use-after-free memory vulnerabilities found in components such as Compositing, Views, Skia, and Ozone. If successfully exploited, these flaws could allow attackers to execute malicious code through specially crafted web content.

The update also fixes:

• Insufficient input validation in the Dawn and ANGLE graphics libraries.
• A race condition in Chrome’s Updater that could lead to unexpected behavior during software updates.

These vulnerabilities could potentially be used to bypass security protections, escalate privileges, or compromise browser stability.

Additional High-Severity Fixes

Beyond the critical issues, Chrome 151 addresses numerous high-severity vulnerabilities affecting components such as:

• V8 JavaScript Engine
• Navigation
• QUIC
• Audio and Media
• WebGL
• Downloads

Many of these flaws involve memory corruption issues, including out-of-bounds reads and writes, integer overflows, and type confusion vulnerabilities, all of which could increase the risk of remote code execution.

Medium and Low-Severity Improvements

The update also includes fixes for medium-severity vulnerabilities affecting features such as Autofill, DevTools, Extensions, Password Manager, WebXR, WebView, and ANGLE.

In addition, Google resolved several lower-severity issues impacting Enterprise features, Bluetooth, NFC, Google Lens, Settings, Picture-in-Picture, and AI-related functionality. While these vulnerabilities are less severe individually, addressing them helps strengthen Chrome’s overall security.

Update Chrome as Soon as Possible

Google credited both its internal security teams and external researchers for identifying many of the vulnerabilities before they could be exploited.

With 370 security fixes, including seven critical vulnerabilities, this release highlights the importance of keeping web browsers up to date. Users and organizations should install Chrome 151 as soon as it becomes available to reduce the risk of attacks targeting known security flaws.

The post Google Chrome 151 Patches Critical Vulnerabilities appeared first on First Hackers News.