Search This Blog
Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Load More
Trendy Right Now

Popular Posts

Showing posts with label Cyber Security. Show all posts
malicious extensions
Chrome Extensions Cookies Cyber Security cybersecurity risks Google Chrome security malicious extensions

Malicious Chrome Extensions Target Enterprise HR and ERP Platforms to Steal Credentials

 

One after another, suspicious Chrome add-ons began appearing under false pretenses - each masquerading as helpful utilities. These were pulled from public view only after Socket, a cybersecurity group, traced them back to a single pattern of abuse. Instead of boosting efficiency, they harvested data from corporate systems like Workday, NetSuite, and SAP SuccessFactors. Installation counts climbed past 2,300 across five distinct apps before takedown. Behind the scenes, threat actors leveraged legitimate-looking interfaces to gain access where it mattered most. 

One investigation found that certain browser add-ons aimed to breach corporate systems, either by capturing login details or disrupting protective measures. Though appearing under distinct titles and author profiles, these tools carried matching coding patterns, operational frameworks, and selection methods - pointing to coordination behind their release. A person using the handle databycloud1104 was linked to four of them; another version emerged through a separate label called Software Access. 

Appearing alongside standard business applications, these extensions asked for permissions typical of corporate tools. One moment they promised better control over company accounts, the next they emphasized locking down admin functions. Positioned as productivity aids, several highlighted dashboard interfaces meant to streamline operations across teams. Instead of standing out, their behavior mirrored genuine enterprise solutions. Claiming to boost efficiency or tighten security, each framed its purpose around workplace demands. Not every feature list matched actual functionality, yet on the surface everything seemed aligned with professional needs. 

Yet the investigation revealed every extension hid its actual operations. Although privacy notices were present, they omitted details about gathering user data, retrieving login information, or tracking admin actions. Without visibility, these tools carried out harmful behaviors - such as stealing authentication cookies, altering webpage elements, or taking over active sessions - all while appearing legitimate. What seemed harmless operated differently beneath the surface. 

Repeated extraction of authentication cookies called "__session" occurred across multiple extensions. Despite user logout actions, those credentials kept reaching external servers controlled by attackers. Access to corporate systems remained uninterrupted due to timed transmissions. Traditional sign-in protections failed because live session data was continuously harvested elsewhere. 

Notably, two add-ons - Tool Access 11 and Data By Cloud 2 - took more aggressive steps. Instead of merely monitoring, they interfered directly with key security areas in Workday. Through recognition of page titles, these tools erased information or rerouted admins before reaching control panels. Pages related to login rules appeared blank or led elsewhere. Controls involving active sessions faced similar disruptions. Even IP-based safeguards vanished unexpectedly. Managing passwords became problematic under their influence. Deactivating compromised accounts grew harder. Audit trails for suspicious activity disappeared without notice. As a result, teams lost vital ground when trying to spot intrusions or contain damage. 

What stood out was the Software Access extension’s ability to handle cookies in both directions. Not only did it take cookies from users, but also inserted ones provided by attackers straight into browsers. Because of this, unauthorized individuals gained access to active sessions - no login details or extra verification steps required. The outcome? Full control over corporate accounts within moments. 

Even with few users impacted, Socket highlighted how compromised business logins might enable wider intrusions - such as spreading ransomware or extracting major datasets. After the discovery, the company alerted Google; soon after, the malicious add-ons vanished from the Chrome Web Store. Those who downloaded them should inform internal security staff while resetting access codes across exposed systems to reduce exposure. Though limited in reach, the breach carries serious downstream implications if left unchecked.
Read more »
Innovation
AI Deepfakes Artificial Intelligence Cyber Security Cyber Threats Financial Data Innovation

Why Cybersecurity Threats in 2026 Will Be Harder to See, Faster to Spread, And Easier to Believe

 


The approach to cybersecurity in 2026 will be shaped not only by technological innovation but also by how deeply digital systems are embedded in everyday life. As cloud services, artificial intelligence tools, connected devices, and online communication platforms become routine, they also expand the surface area for cyber exploitation.

Cyber threats are no longer limited to technical breaches behind the scenes. They increasingly influence what people believe, how they behave online, and which systems they trust. While some risks are still emerging, others are already circulating quietly through commonly used apps, services, and platforms, often without users realizing it.

One major concern is the growing concentration of internet infrastructure. A substantial portion of websites and digital services now depend on a limited number of cloud providers, content delivery systems, and workplace tools. This level of uniformity makes the internet more efficient but also more fragile. When many platforms rely on the same backbone, a single disruption, vulnerability, or attack can trigger widespread consequences across millions of users at once. What was once a diverse digital ecosystem has gradually shifted toward standardization, making large-scale failures easier to exploit.

Another escalating risk is the spread of misleading narratives about online safety. Across social media platforms, discussion forums, and live-streaming environments, basic cybersecurity practices are increasingly mocked or dismissed. Advice related to privacy protection, secure passwords, or cautious digital behavior is often portrayed as unnecessary or exaggerated. This cultural shift creates ideal conditions for cybercrime. When users are encouraged to ignore protective habits, attackers face less resistance. In some cases, misleading content is actively promoted to weaken public awareness and normalize risky behavior.

Artificial intelligence is further accelerating cyber threats. AI-driven tools now allow attackers to automate tasks that once required advanced expertise, including scanning for vulnerabilities and crafting convincing phishing messages. At the same time, many users store sensitive conversations and information within browsers or AI-powered tools, often unaware that this data may be accessible to malware. As automated systems evolve, cyberattacks are becoming faster, more adaptive, and more difficult to detect or interrupt.

Trust itself has become a central target. Technologies such as voice cloning, deepfake media, and synthetic digital identities enable criminals to impersonate real individuals or create believable fake personas. These identities can bypass verification systems, open accounts, and commit fraud over long periods before being detected. As a result, confidence in digital interactions, platforms, and identity checks continues to decline.

Future computing capabilities are already influencing present-day cyber strategies. Even though advanced quantum-based attacks are not yet practical, some threat actors are collecting encrypted data now with the intention of decrypting it later. This approach puts long-term personal, financial, and institutional data at risk and underlines the need for stronger, future-ready security planning.

As digital and physical systems become increasingly interconnected, cybersecurity in 2026 will extend beyond software and hardware defenses. It will require stronger digital awareness, better judgment, and a broader understanding of how technology shapes risk in everyday life.

Read more »
Malwares
Cyber Security GootLoader Malicious Codes Malicious Files Malware attacks Malwares

GootLoader Malware Uses Malformed ZIP Archives to Evade Detection

 

A fresh tactic has emerged among cybercriminals using GootLoader, a JavaScript-driven malware installer. Instead of standard compression, they now distribute broken ZIP files designed to slip past digital defenses. These flawed archives exploit differences across decompression programs - some fail to process them, others do so partially. This mismatch lets malicious code stay concealed during scans yet run normally when opened by users. Findings detailed by Expel show that inconsistent parsing logic in software plays right into attacker hands. Hidden scripts activate only when handled by specific tools found on typical machines. 

Starting with a strange structure, these harmful ZIP files combine around 500 to 1,000 smaller archives into one large package. Because of this layered setup, standard programs like WinRAR or 7-Zip cannot properly read them - tools often relied on during malware checks. Due to the confusion they create, automatic detection systems frequently skip examining what's inside. Yet, when opened through Windows’ own built-in decompression feature, the file works without issue. 

That smooth operation lets victims unknowingly unpack dangerous content. Since 2020, GootLoader has maintained a presence among cyber threats, primarily spreading via manipulated search results and deceptive online ads. People looking for official forms or corporate paperwork may unknowingly land on hacked WordPress sites offering infected files. These corrupted archives, once opened, trigger the payload delivery mechanism embedded within the software. Acting as a gateway tool, it paves the way for additional harmful programs - ransomware being one frequent outcome. 

The chain of infection begins quietly, escalating quickly under the radar. By late 2025, Expel researchers noticed subtle upgrades, showing how the attack method keeps shifting. Instead of just stacking archives, hackers shorten key metadata inside ZIP structures - especially tampering with the end of central directory entries. That tweak triggers failures in numerous analysis programs, yet files still open in Windows Explorer. 

Inside the package, unimportant sections get scrambled too, throwing off predictable reading patterns and making automated inspection harder. Researchers refer to this method as "hashbusting," delivering a distinct ZIP file to each target. Every time someone downloads it, differences in the archive's layout and data prevent standard hash checks from working. Even the JavaScript inside changes form with each instance. Detection systems relying on repeated patterns struggle as a result. 

 What makes the delivery hard to catch lies in its method. Rather than sending a typical ZIP archive, attackers transmit the malicious code as an XOR-encrypted flow of data, rebuilt only after reaching the target's browser. It grows by adding copies of itself over and over, expanding until it meets a specific volume - this skirts detection meant for compressed files. After launch, the script runs using built-in Windows tools, skipping any need to unpack completely, so the attack unfolds without drawing attention. 

Once active, it stays on the machine by placing shortcuts into the Windows Startup directory - then triggers further scripts through native utilities like cscript or PowerShell. From there, data collection begins: details about the system get pulled and sent back to distant servers that control the attack, setting up what comes next without delay. 

Although often overlooked, limiting access to built-in tools such as wscript.exe helps block common attack paths. Instead of running scripts automatically, setting systems to display code in basic viewers adds another layer of protection. As seen with GootLoader’s shifts over time, attackers now twist everyday OS functions into stealthy weapons, staying active even when defenses improve.
Read more »
URL
Artificial Intelligence Cyber Security Data Exfiltration Microsoft Copilot Reprompt URL

Security Researchers Warn of ‘Reprompt’ Flaw That Turns AI Assistants Into Silent Data Leaks

 



Cybersecurity researchers have revealed a newly identified attack technique that shows how artificial intelligence chatbots can be manipulated to leak sensitive information with minimal user involvement. The method, known as Reprompt, demonstrates how attackers could extract data from AI assistants such as Microsoft Copilot through a single click on a legitimate-looking link, while bypassing standard enterprise security protections.

According to researchers, the attack requires no malicious software, plugins, or continued interaction. Once a user clicks the link, the attacker can retain control of the chatbot session even if the chat window is closed, allowing information to be quietly transmitted without the user’s awareness.

The issue was disclosed responsibly, and Microsoft has since addressed the vulnerability. The company confirmed that enterprise users of Microsoft 365 Copilot are not affected.

At a technical level, Reprompt relies on a chain of design weaknesses. Attackers first embed instructions into a Copilot web link using a standard query parameter. These instructions are crafted to bypass safeguards that are designed to prevent direct data exposure by exploiting the fact that certain protections apply only to the initial request. From there, the attacker can trigger a continuous exchange between Copilot and an external server, enabling hidden and ongoing data extraction.

In a realistic scenario, a target might receive an email containing what appears to be a legitimate Copilot link. Clicking it would cause Copilot to execute instructions embedded in the URL. The attacker could then repeatedly issue follow-up commands remotely, prompting the chatbot to summarize recently accessed files, infer personal details, or reveal contextual information. Because these later instructions are delivered dynamically, it becomes difficult to determine what data is being accessed by examining the original prompt alone.

Researchers note that this effectively turns Copilot into an invisible channel for data exfiltration, without requiring user-entered prompts, extensions, or system connectors. The underlying issue reflects a broader limitation in large language models: their inability to reliably distinguish between trusted user instructions and commands embedded in untrusted data, enabling indirect prompt injection attacks.

The Reprompt disclosure coincides with the identification of multiple other techniques targeting AI-powered tools. Some attacks exploit chatbot connections to third-party applications, enabling zero-interaction data leaks or long-term persistence by injecting instructions into AI memory. Others abuse confirmation prompts, turning human oversight mechanisms into attack vectors, particularly in development environments.

Researchers have also shown how hidden instructions can be planted in shared documents, calendar invites, or emails to extract corporate data, and how AI browsers can be manipulated to bypass built-in prompt injection defenses. Beyond software, hardware-level risks have been identified, where attackers with server access may infer sensitive information by observing timing patterns in machine learning accelerators.

Additional findings include abuses of trusted AI communication protocols to drain computing resources, trigger hidden tool actions, or inject persistent behavior, as well as spreadsheet-based attacks that generate unsafe formulas capable of exporting user data. In some cases, attackers could manipulate AI development platforms to alter spending controls or leak access credentials, enabling stealthy financial abuse.

Taken together, the research underlines that prompt injection remains a persistent and evolving risk. Experts recommend layered security defenses, limiting AI privileges, and restricting access to sensitive systems. Users are also advised to avoid clicking unsolicited AI-related links and to be cautious about sharing personal or confidential information in chatbot conversations.

As AI systems gain broader access to corporate data and greater autonomy, researchers warn that the potential impact of a single vulnerability increases substantially, underscoring the need for careful deployment, continuous monitoring, and ongoing security research.


Read more »
Websites
Credit Card Cyber Security Financial Data Web Skimming Websites

Ongoing Web Skimming Operation Quietly Harvests Payment Data From Online Stores

 



Cybersecurity analysts have identified a sophisticated web skimming operation that has been running continuously since early 2022, silently targeting online checkout systems. The campaign focuses on stealing payment card information and is believed to affect businesses that rely on globally used card networks.

Web skimming is a type of cyberattack where criminals tamper with legitimate shopping websites rather than attacking customers directly. By inserting malicious code into payment pages, attackers are able to intercept sensitive information at the exact moment a customer attempts to complete a purchase. Because the website itself appears normal, victims are usually unaware their data has been compromised.

This technique is commonly associated with Magecart-style attacks. While Magecart initially referred to groups exploiting Magento-based websites, the term now broadly describes any client-side attack that captures payment data through infected checkout pages across multiple platforms.

The operation was uncovered during an investigation into a suspicious domain hosting malicious scripts. This domain was linked to infrastructure previously associated with a bulletproof hosting provider that had faced international sanctions. Researchers found that the attackers were using this domain to distribute heavily concealed JavaScript files that were loaded directly by e-commerce websites.

Once active, the malicious script continuously monitors user activity on the payment page. It is programmed to detect whether a website administrator is currently logged in by checking for specific indicators commonly found on WordPress sites. If such indicators are present, the script automatically deletes itself, reducing the risk of detection during maintenance or inspection.

The attack becomes particularly deceptive when certain payment options are selected. In these cases, the malicious code creates a fake payment form that visually replaces the legitimate one. Customers unknowingly enter their card number, expiration date, and security code into this fraudulent interface. After the information is captured, the website displays a generic payment error, making it appear as though the transaction failed due to a simple mistake.

In addition to financial data, the attackers collect personal details such as names, contact numbers, email addresses, and delivery information. This data is sent to an external server controlled by the attackers using standard web communication methods. Once the transfer is complete, the fake form is removed, the real payment form is restored, and the script marks the victim as already compromised to avoid repeating the attack.

Researchers noted that the operation reflects an advanced understanding of website behavior, especially within WordPress-based environments. By exploiting both technical features and user trust, the attackers have managed to sustain this campaign for years without drawing widespread attention.

This discovery reinforces the importance of continuous website monitoring and script validation for businesses, as well as cautious online shopping practices for consumers.

Read more »
Organization security
Corporate Cyber Security data security EOCC Organization security

EOCC Hit by Security Breach Due to Contractor's Unauthorised Access


The Equal Employment Opportunity Commission (EOCC) was hit by an internal security data breach that happened last year. The incident involved a contractor's employees exploiting sensitive data in an agency's systems. 

About the breach

The breach happened in EEOC's Public Portal system where unauthorized access of agency data may have disclosed personal data in logs given to agency by the public. “Staff employed by the contractor, who had privileged access to EEOC systems, were able to handle data in an unauthorized (UA) and prohibited manner in early 2025,” reads the EEOC email notification sent by data security office. 

The email said that the review suggested personally identifiable information (PII) may have been leaked, depending on the individual. The exposed information may contain names, contact and other data. The review of is still ongoing while EOCC works with the law enforcement. 

EOCC has asked individuals to review their financial accounts for any malicious activity and has also asked portal users to reset their passwords. 

Contracting data indicates that EEOC had a contract with Opexus, a company that provides case management software solutions to the federal government.

 Prevention measures 

Open spokesperson confirmed this and said EEOC and Opex “took immediate action when we learned of this activity, and we continue to support investigative and law enforcement efforts into these individuals’ conduct, which is under active prosecution in the Federal Court of the Eastern District of Virginia.” 

Talking about the role of employees in the breach, the spokesperson added that “While the individuals responsible met applicable seven-year background check requirements consistent with prevailing government and industry standards at the time of hire, this incident made clear that personnel screening alone is not sufficient." 

The second Trump administration's efforts to prevent claimed “illegal discrimination” driven by diversity, equity, and inclusion programs, which over the past year have been examined and demolished at almost every level of the federal government, centre on the EEOC. 

Large private companies all throughout the nation have been affected by the developments. In an X post this month, EEOC chairwoman Andrea Lucas asked white men if they had experienced racial or sexual discrimination at work and urged them to report their experiences to the organization "as soon as possible.”

Read more »
Malware Threat
AsyncRAT AsyncRAT attack Cloudfare Cyber Phishing Cyber Security Cyberattacks cybersecurity risks Malware Threat

AsyncRAT Campaign Abuses Cloudflare Services to Hide Malware Operations

 

Cybercriminals distributing the AsyncRAT remote access trojan are exploiting Cloudflare’s free-tier services and TryCloudflare tunneling domains to conceal malicious infrastructure behind widely trusted platforms. By hosting WebDAV servers through Cloudflare, attackers are able to mask command-and-control activity, making detection significantly more difficult for conventional security tools that often whitelist Cloudflare traffic. 

The campaign typically begins with phishing emails that contain Dropbox links. These links deliver files using double extensions, such as .pdf.url, which are designed to mislead recipients into believing they are opening legitimate documents. When the files are opened, victims unknowingly download multi-stage scripts from TryCloudflare domains. At the same time, a genuine PDF document is displayed to reduce suspicion and delay user awareness of malicious activity. 

A notable aspect of this operation is the attackers’ use of legitimate software sources. The malware chain includes downloading official Python distributions directly from Python.org. Once installed, a full Python environment is set up on the compromised system. This environment is then leveraged to execute advanced code injection techniques, specifically targeting the Windows explorer.exe process, allowing the malware to run stealthily within a trusted system component. 

To maintain long-term access, the attackers rely on multiple persistence mechanisms. These include placing scripts such as ahke.bat and olsm.bat in Windows startup folders so they automatically execute when a user logs in. The campaign also uses WebDAV mounting to sustain communication with command-and-control servers hosted through Cloudflare tunnels. 

The threat actors heavily employ so-called “living-off-the-land” techniques, abusing built-in Windows tools such as PowerShell, Windows Script Host, and other native utilities. By blending malicious behavior with legitimate system operations, the attackers further complicate detection and analysis, as their activity closely resembles normal administrative actions. 

According to research cited by Trend Micro, the use of Cloudflare’s infrastructure creates a significant blind spot for many security solutions. Domains containing “trycloudflare.com” often appear trustworthy, allowing AsyncRAT payloads to be delivered without triggering immediate alerts. This abuse of reputable services highlights how attackers increasingly rely on legitimate platforms to scale operations and evade defenses. 

Security researchers warn that although known malicious repositories and infrastructure may be taken down, similar campaigns are likely to reappear using new domains and delivery methods. Monitoring WebDAV connections, scrutinizing traffic involving TryCloudflare domains, and closely analyzing phishing attachments remain critical steps in identifying and mitigating AsyncRAT infections.
Read more »
Sensitive data
AI Agent ai application Cyber Security LangChain Sensitive data

LangChain Security Issue Puts AI Application Data at Risk

 



A critical security vulnerability has been identified in LangChain’s core library that could allow attackers to extract sensitive system data from artificial intelligence applications. The flaw, tracked as CVE-2025-68664, affects how the framework processes and reconstructs internal data, creating serious risks for organizations relying on AI-driven workflows.

LangChain is a widely adopted framework used to build applications powered by large language models, including chatbots, automation tools, and AI agents. Due to its extensive use across the AI ecosystem, security weaknesses within its core components can have widespread consequences.

The issue stems from how LangChain handles serialization and deserialization. These processes convert data into a transferable format and then rebuild it for use by the application. In this case, two core functions failed to properly safeguard user-controlled data that included a reserved internal marker used by LangChain to identify trusted objects. As a result, untrusted input could be mistakenly treated as legitimate system data.

This weakness becomes particularly dangerous when AI-generated outputs or manipulated prompts influence metadata fields used during logging, event streaming, or caching. When such data passes through repeated serialization and deserialization cycles, the system may unknowingly reconstruct malicious objects. This behavior falls under a known security category involving unsafe deserialization and has been rated critical, with a severity score of 9.3.

In practical terms, attackers could craft inputs that cause AI agents to leak environment variables, which often store highly sensitive information such as access tokens, API keys, and internal configuration secrets. In more advanced scenarios, specific approved components could be abused to transmit this data outward, including through unauthorized network requests. Certain templating features may further increase risk if invoked after unsafe deserialization, potentially opening paths toward code execution.

The vulnerability was discovered during security reviews focused on AI trust boundaries, where the researcher traced how untrusted data moved through internal processing paths. After responsible disclosure in early December 2025, the LangChain team acknowledged the issue and released security updates later that month.

The patched versions introduce stricter handling of internal object markers and disable automatic resolution of environment secrets by default, a feature that was previously enabled and contributed to the exposure risk. Developers are strongly advised to upgrade immediately and review related dependencies that interact with LangChain-core.

Security experts stress that AI outputs should always be treated as untrusted input. Organizations are urged to audit logging, streaming, and caching mechanisms, limit deserialization wherever possible, and avoid exposing secrets unless inputs are fully validated. A similar vulnerability identified in LangChain’s JavaScript ecosystem accentuates broader security challenges as AI frameworks become more interconnected.

As AI adoption accelerates, maintaining strict data boundaries and secure design practices is essential to protecting both systems and users from newly developing threats.

Read more »
Subscribe to: Comments (Atom)