Search This Blog

Powered by Blogger.

Blog Archive

Labels

Footer About

Footer About

Labels

Showing posts with label URL. Show all posts

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.


Cybercriminals Target AI Enthusiasts with Fake Websites to Spread Malware

 


Cyber attackers are now using people’s growing interest in artificial intelligence (AI) to distribute harmful software. A recent investigation has uncovered that cybercriminals are building fake websites designed to appear at the top of Google search results for popular AI tools. These deceptive sites are part of a strategy known as SEO poisoning, where attackers manipulate search engine algorithms to increase the visibility of malicious web pages.

Once users click on these links believing they’re accessing legitimate AI platforms they’re silently redirected to dangerous websites where malware is secretly downloaded onto their systems. The websites use layers of code and redirection to hide the true intent from users and security software.

According to researchers, the malware being delivered includes infostealers— a type of software that quietly gathers personal and system data from a user’s device. These can include saved passwords, browser activity, system information, and more. One type of malware even installs browser extensions designed to steal cryptocurrency.

What makes these attacks harder to detect is the attackers' use of trusted platforms. For example, the malicious code is sometimes hosted on widely used cloud services, making it seem like normal website content. This helps the attackers avoid detection by antivirus tools and security analysts.

The way these attacks work is fairly complex. When someone visits one of the fake AI websites, their browser is immediately triggered to run hidden JavaScript. This script gathers information about the visitor’s browser, encrypts it, and sends it to a server controlled by the attacker. Based on this information, the server then redirects the user to a second website. That second site checks details like the visitor’s IP address and location to decide whether to proceed with delivering the final malicious file.

This final step often results in the download of harmful software that invades the victim’s system and begins stealing data or installing other malicious tools.

These attacks are part of a growing trend where the popularity of new technologies, such as AI chatbots is being exploited by cybercriminals for fraudulent purposes. Similar tactics have been observed in the past, including misleading users with fake tools and promoting harmful applications through hijacked social media pages.

As AI tools become more common, users should remain alert while searching for or downloading anything related to them. Even websites that appear high in search engine results can be dangerous if not verified properly.

To stay safe, avoid clicking on unfamiliar links, especially when looking for AI services. Always download tools from official sources, and double-check website URLs. Staying cautious and informed is one of the best ways to avoid falling victim to these evolving online threats.

URL Scams Everywhere? These Steps Will Help You Stay Safe

Scams Everywhere? These Steps Will Help You Stay Safe

Scam links are difficult to spot, but it has become an everyday issue for internet users who accidentally click on malicious URLs that are part of a phishing attack. Most fake links include standard “https” encryption and domains similar to real websites. Phishing and spoofing scams caused over $70 million in losses for victims in 2024 says FBI’s Internet Crime Complaint Center. 

When users click on a scam link, they might suffer monetary losses, and worse, give up private info such as name and credit card details to scammers, they may also accidentally install malware on their device. 

How to spot scam link

They are generally found in text messages and emails sent by scammers, designed to trick us into downloading malware or bringing us to a scam website to steal our personal identifying information. A few examples include gold bars, employment, and unpaid toll scams. Scammers send these links to the masses— with the help of AI these days. Since a lot of users fall victim to phishing scams every year,  scammers haven’t changed their attack tactics over the years.

How to avoid scam link

Always check the URL

These days, smartphones try to block scam links, so scammers have adapted making links that escape detection. Users are advised to look for typos-quatting, a technique that uses spelling mistakes. For eg: 'applle' instead of 'apple'. 

Be cautious of URLs you visit regularly

Most brands don’t change their domain names. If you find the domain name is different in the URL, it is a fake link. 

Watch out for short links

Shortlists are generally found on social media and texts. Experts say there is no way to determine the authenticity of a shortened URL, advising users to not open them. Instead, users should check the language for any suspicious signs. 

How do victims receive scam links?

Text scams

These don’t need website links, they are sent via phone numbers. Users accidentally click on a malicious phone number thinking it is their bank or someone important. Experts suggest not to interact with unknown phone numbers. 

Email

The most popular means to send scam links is via e-mail, resulting in the biggest monetary losses. To stay safe, users can copy the link in their notepad first and inspect it before opening it. 

QR code scams

Malicious QR codes have become common in public avenues, from restaurants to parking stands. Scammers embed fake codes over real ones or fill them with phishing emails that redirect to fake sites or malware downloads. 

DMs on social media

Scammers pretend to be someone you know, they may fake a medical emergency and demand you for money to help them. Always call the person to cross-check the identity before giving money, opening a link, or revealing any personal information. 

How to Check If a Downloaded File Is Safe to Use

 


It is no longer a secret that downloading software is becoming an integral part of everyday computing in today’s digitally based environment. It is used to enhance productivity, explore new tools, and stay connected to an ever-increasing online world, all of which are aided by downloads of software. While instant downloads have many advantages, if they are not approached with due diligence, they can also pose significant risks. 

A variety of harmful software, including malware, spyware, and adware, can be easily embedded into seemingly harmless files, potentially compromising personal information or system functionality. Given this, users need to take a cautious and informed approach before they execute any downloaded file. 

By following a few simple steps to verify a file’s safety, for example, scanning it for antivirus, and signing it with a digital signature, users can greatly reduce their vulnerability to cybersecurity risks. 

As digital threats continue to evolve, awareness and prevention remain the best defences for a constantly evolving cyber environment. While downloading files from the internet is now part of current daily lives, it is not without its risks. Cybercriminals often take advantage of this habit by disguising malicious software, like viruses, trojans, ransomware, and a wide variety of other forms of malware, as legitimate software. 

The threats are often disguised as harmless files, making it easy for the uninitiated to become victims of data loss or security breaches. This is why it is imperative to use caution when downloading any content, regardless of the source, regardless of whether the source seems trustworthy. The risk of infection can be significantly reduced by practising due diligence by scanning files using antivirus software, checking for digital signatures, and avoiding unknown or suspicious links when it comes to downloading files. 

With the ever-evolving digital threat landscape, users must take precautions about file safety, not just as a recommendation, but as a necessity. Users across the globe are increasingly concerned about the risk of downloading malicious software unintentionally from the internet. It is possible to install malicious programs on a computer system just by clicking a single careless button. 

A malicious program could compromise the integrity of the system, take sensitive data, or render a computer inoperable. As a result of SonicWall's Cyber Threat Report 2021, there were more than 5.6 billion malware attacks recorded in 2020 alone, a staggering figure that indicates how persistent this threat has become. 

A malware infection is usually caused by deceptive email attachments, compromised websites, and software downloads that appear legitimate but are laced with hidden dangers, resulting in the infection of a device. As a result, many users unknowingly expose themselves to such risks when they install a file or application that they believe is safe and secure. As a result, it highlights the importance of being vigilant and informed when it comes to navigating the digital world. Anyone who wants to protect their digital environment must understand how malware spreads, adopt proactive safety habits, and become aware of the dangers lurking within downloadable files.

For organisations to strengthen their cybersecurity protocols, it is imperative to have a thorough understanding of the hidden threats lurking within downloadable files. A fairly common infection vector is malicious email attachments that are sent as part of an email. There is a common practice among cybercriminals of using deceptive emails to distribute infected files disguised as regular documents, such as invoices, reports, or internal memos, that contain infected files. It has been shown that these attachments can unleash email-based viruses which will infiltrate entire company networks and spread quickly, leading to widespread disruption. There is also a threat vector that resides within seemingly harmless documents from Microsoft Office. 

Word or Excel documents, for example, may contain malicious macros—automated scripts embedded within them. When an unsuspecting recipient enables macros, these scripts silently execute, causing the system to be compromised with malware. These types of attacks are especially dangerous because they appear to be standard business communication when they are, in fact, very dangerous. 

Compressed files such as .zip and .rar also pose a significant threat. Often, threat actors hide harmful executable files within these archives, making it more difficult for them to be detected. Once those files are extracted and executed, they can instantly infect a device, granting unauthorized access, or causing further damage to the network infrastructure. 

Given that these threats are becoming increasingly sophisticated and subtlebusinesses must develop proactive strategies that can prevent them from becoming infected in the first place. An organization might be able to prevent malicious software from entering its organisation by implementing comprehensive employee training programs, strict file filtering policies, advanced threat detection tools, and regular updates to software. 

The prevention of malicious software begins with awareness and continues through rigorous cybersecurity practices and disciplined digital hygiene. There is a potential security risk associated with every file that user download from the internet, whether it is a file attached to an email, a multimedia file, or something that appears harmless like a screen saver. It is possible for familiar sources to unknowingly transmit compromised files, which is why vigilance is essential in every digital interaction. 

Here are a few critical practices that need to be followed to protect both personal devices and organisational networks. To greatly reduce the possibility of infection with harmful software, it is imperative to exercise digital caution and apply sound judgment by avoiding downloads from unknown or suspicious sources. Users are significantly less likely to become infected with dangerous software. When users initiate a download, they should use a reputable website that has a secure (HTTPS) connection and has a well-known domain name. 

Users can prevent fraud by checking the URL bar of the site to ensure its legitimacy. Moreover, fraudulent emails continue to be a very common vehicle for distributing malware. Links and attachments within unsolicited or unexpected messages should never be opened without verifying that the source is genuine. If users encounter suspicious pop-ups or warnings while browsing, they would be wise to close them by clicking the close (X) button in the browser rather than engaging with them. 

A second method of protecting against malware is to save files on people's devices before opening them, which will allow their antivirus software to scan them and alert them to any potential threats that may exist. In addition to verifying the file extension, reading user reviews and comments can provide valuable insights, as previous users may have already reported security issues or hidden dangers.

Media files, for example, should never be delivered in executable (.exe) format, because this indicates malicious intent. Although these practices are simple in nature, they nonetheless serve as a powerful means of avoiding the growing threat of a complex and constantly evolving digital environment. 

Importance of Robust Antivirus and Antimalware Software 


Luigi Oppido, a computer expert, emphasised the importance of installing reputable antispyware, antivirus and antispyware programs such as Norton, AVG, Malwarebytes, or Avast. These programs provide an important line of defence by actively scanning files as soon as they are downloaded, which provides a vital line of defence by identifying and blocking malicious software before it reaches users' computers. Antivirus applications are often integrated into operating systems, which should be enabled and monitored for any security alerts to make sure they do not get infected. 

Download from Trusted Sources 


It is important to note that files obtained exclusively from official websites of established companies, like Microsoft, are much less likely to have any malware attached to them. In contrast, downloading files from less well known or unreliable websites poses a higher threat. In addition to enhancing security, using official digital distribution platforms such as Microsoft Store or Apple App Store adds another layer of protection since these platforms thoroughly vet software before listing it. 

Verify Website Authenticity


As a result of cybercriminals creating spoofed websites using subtle variations in the domain names, users can often be deceived by spoofed sites (e.g., “microsoft.co” rather than “microsoft.com”). As a guide, users should look for signs of a trustworthy site, including a professional site design, a lack of excessive pop-ups or spam links, and the presence of SSL/TLS certificates, which can be recognised by the “https” and padlock icon on the browser. 

Awareness of Download Context 


A significant portion of the risk associated with downloading a file is determined by the source of the download. Files from dubious places, like torrent sites or adult content platforms, are often highly dangerous, and often contain malware or viruses. Files that resemble official software or originate from reputable companies are generally less dangerous.

Recognise Browser and System Warnings

It is important for users to heed warnings sent by modern browsers and antivirus programs when they are interested in downloading suspicious websites or potentially dangerous files. They must acknowledge these warnings and avoid proceeding with questionable downloads.

Check User Feedback and File Reputation


Reviews and comments left by users, whether on the hosting website or independent forums such as Reddit and Quora, can offer insights into the safety of a download. A positive reaction from multiple users will typically indicate a lower risk of malware infection. 

File Size Considerations


Several clues can be provided by the file size of a file. Usually, the size of a file is an indication of its legitimacy. An unusually small file may contain incomplete data or disguised malware. An unexpectedly large file may carry unwanted or harmful extras along with its intended purpose. 

Caution with Executable and Archive Files


It is common for malware to manifest itself in executable files (e.g., “.exe,” “.bat,” “.msi,” “.scr”) that were sourced from unknown locations. Hackers often use double extensions such as “.gif.exe” in order to trick consumers into executing harmful software. People using devices like laptops, computers, or mobiles must verify the source and digital signature of the executable file before opening it, since it grants an individual extensive control over the system. 

Digital Signatures and Licensing


Whenever users are running software on Windows, digital signatures and license warnings serve as indicators of authenticity. There is no guarantee that every executable is safe, no guarantee that every executable is intended to do harm. However, these factors can guide risk assessments before the installation of software is performed. 

The temptation to bypass security alerts, such as those that appear after a Windows update or warn that i file is potentially dangerous, arises whenever software is installed, and in the rush to do so, security warnings can be easily dismissed or disabled. However, these alerts serve a crucial function in protecting systems against potential threats. 

With Windows SmartScreen and other similar security mechanisms, users get more than just traditional antivirus software; they look at file reputations and behavioural patterns, which can often allow them to detect malware that conventional signature-based scanners may miss. As a precautionary measure, rather than switching off these protections, it is prudent to use such alerts as an opportunity to assess the file's safety using well-established verification methods rather than turning them off.

A major point to remember is that legitimate software rarely triggers multiple security warnings; encountering several warnings should be considered a clear red flag, indicating that the file may pose serious risks. To prevent infections and ensure the integrity of computer systems, one must maintain constant vigilance and respect these security layers.

2FA Under Attack as Astaroth Phishing Kit Spreads

 


Astaroth is the latest phishing tool discovered by cybercriminals. It has advanced capabilities that allow it to circumvent security measures such as two-factor authentication (2FA) when used against it. In January 2025, Astaroth made its public debut across multiple platforms, including Gmail, Yahoo, and Office 365, with sophisticated technologies such as session hijacking and real-time credentials interceptions, which compromise user accounts across multiple platforms. 

SlashNext researchers claim Astaroth makes use of a reverse proxy called an evilginx-style proxy to place itself between legitimate login pages and users. As a result, the tool is capable of intercepting and capturing sensitive credentials, such as usernames, passwords, 2FA tokens, and session cookies, without triggering security alerts, thereby making the tool effective. 

It has been demonstrated that attackers who have obtained these session cookies will be able to hijack authenticated sessions, bypass additional security protocols, and gain unauthorized access to user accounts once they have acquired these cookies. Astaroth demonstrates the evolution of cyber threats and the sophistication of phishing techniques that compromise online security. This development highlights how cybercriminals have been evolving their methods of phishing over the years.

Clearly, Astaroth highlights how cybercriminals' tactics have evolved over the last decade, as phishing has evolved into a lucrative business. The sophistication of sophisticated attacks has now reached a point where it is now marketed like commercial software products, with regular updates, customer support, and testing guarantees attached to them. 

The attacker can intercept real-time credentials and use reverse proxy techniques in order to hijack authenticated sessions in order to bypass even the most robust phishing defences, such as Multi Factor Authentication (MFA), which are designed to protect against phishing attacks. Due to the widespread availability of phishing kits such as Astaroth, which significantly reduces the barrier to entry, less experienced cybercriminals are now capable of conducting highly effective attacks given that the barriers to entry have been significantly lowered. 

The key to mitigating these threats is to adopt a comprehensive, multilayered security strategy that is both comprehensive and multifaceted. It must have a password manager, endpoint security controls, real-time threat monitoring, and ongoing employee training to ensure that employees are aware of cybersecurity threats in real time. 

As an additional consideration, implementing Privillege Access Management (PAM) is equally vital, since it prevents unauthorized access to critical systems, even if login credentials are compromised, through the use of PAM. Business owners remain vulnerable to increasingly sophisticated phishing techniques that can circumvent the traditional defenses of their organisations without appropriate proactive security measures. 

The Astaroth phishing kit has been developed to enable a more effective method of bypassing multi-factor authentication (MFA). By using an evilginx reverse proxy, it intercepts authentication processes in real time as they are happening. By using Astaroth, attackers will be able to steal authenticated sessions and hack them seamlessly with no technical knowledge. Astaroth is different from traditional phishing tools, which capture only static credentials; instead, it dynamically retrieves authorization tokens, 2FA tokens, and session cookies. This tool is a man-in-the-middle attack that renders conventional anti-phishing defenses and multi-factor authentication protections ineffective by acting as an intermediary. 

Discovered by SlashNext Threat Researchers on cybercrime marketplaces, Astaroth is marketed as a tool that can be used easily. It is a 2-in-1 solution that sells for $2000 and includes six months of continuous updates, which includes the newest bypass techniques, as well as pre-purchase testing to demonstrate its effectiveness in real-world attacks if the buyer wants to establish credibility within cybercriminal networks. There is no doubt that the sophistication of phishing kits such as Astaroth, as well as the implementation of behaviour-based authentication, endpoint security controls, and continuous threat monitoring, are critical to organizations in order to defend themselves from these ever-evolving cyber threats that are continually evolving. 

As a means of expanding the company's customer base, Astaroth's developers have publicly revealed the methodologies they use to bypass security measures, such as reCAPTCHA or BotGuard, as a way of demonstrating the kit's effectiveness at circumventing automatic security measures. Cybercriminals in cybercrime forums and underground marketplaces are actively promoting Astaroth among their communities and are primarily distributing it through Telegram, leading to its widespread adoption among cybercriminals world-wide. 

There are several advantages to using these platforms, the most important of which is their accessibility, along with the anonymity they provide. This makes monitoring, tracking, and disrupting the sale and distribution of phishing kits very challenging for law enforcement agencies. There is a particular application known as Telegram which is commonly used by cybercriminals to communicate and to distribute their illicit activities due to its end-to-end encryption, private groups, and minimal oversight. This makes it very difficult for law enforcement to trace illicit activities on Telegram. 

It may not only facilitate the proliferation of Astaroth on the dark web, but also on underground marketplaces - both of which allow threat actors to engage in peer-to-peer transactions without disclosing their identities to each other. The fact that these platforms are decentralized, along with the fact that cryptocurrency payments are used in conjunction with them, adds more layers of protection for cybercriminals, making it even more difficult for authorities to take enforcement action against them. Astaroth continue to be embraced by cybercriminal communities and is lowering the barrier to entry for less-experienced attackers, which in turn is promoting phishing-as-a-service (PhaaS) models which are becoming more prevalent as a consequence. 

Due to the complexities posed by sophisticated phishing kits like Astaroth, security professionals emphasize the need for proactive security measures, which include real-time threat intelligence, endpoint detection, and multi-layered authentication strategies, as well as real-time threat intelligence. Aside from offering custom hosting solutions, Astaroth also offers bulletproof hosting, which will make Astaroth more resilient against legal authorities’ efforts to take down its websites. 

Cybercriminals are able to conduct attacks with minimal disruption in jurisdictions with weak regulatory oversight when using the phishing kit since it operates in jurisdictions that lack regulatory oversight. As a Field CTO of SlashNext, J Stephen Kowski believes that the emergence of Astaroth with regards to authentication is one of the most important implication that could be borne out by the fact that even the most robust authentication systems can be compromised if the attackers obtain the two-factor authentication (2FA) codes and session information during the authentication process in real time. 

Thomas Richards, Principal Consultant and Network and Red Team Practice Director at Black Duck, a Burlington, Massachusetts-based provider of application security solutions, has emphasized the sophistication and severity of the Astaroth phishing kit. According to Richards, this phishing kit demonstrates an advanced level of complexity, making it increasingly difficult for users to identify and avoid such attacks. "Traditional security awareness training often instructs users to recognize phishing attempts by looking for red flags such as suspicious URLs, grammatical errors, or lack of SSL certification. 

However, Astaroth’s highly sophisticated approach significantly reduces these indicators, making detection far more challenging," Richards stated. Furthermore, the infrastructure supporting these attacks is often hosted by providers that do not cooperate with law enforcement agencies, complicating efforts to dismantle these operations. In response to this growing threat, the United States and several European nations have imposed sanctions on countries that provide bulletproof hosting services, which are frequently exploited by cybercriminals to evade legal action. 

Richards advises users to exercise extreme caution when receiving emails that appear to originate from legitimate organizations and contain urgent requests for immediate action. Rather than clicking on embedded links, users should manually navigate to the official website to verify the authenticity of any alerts or account-related issues. This proactive approach is essential in mitigating the risks posed by advanced phishing campaigns like Astaroth. 

Organizations must implement advanced security measures beyond traditional login protections in order to protect themselves from these threats. According to Thomas Richards, a Principal Consultant and Network and Red Team Practice Director for Black Duck, a Burlington-based company that provides applications security solutions, Astaroth's phishing kit is sophisticated and quite severe. As Richards points out, this phishing kit shows a remarkable degree of complexity, which makes it increasingly difficult for users to identify and avoid attacks such as these as they run across them. 

It has always been taught to users during traditional security awareness training to look for red flags, such as suspicious URLs, grammatical errors, or a lack of SSL certification, so they can identify phishing attempts. Although these indicators are largely reduced by Astaroth's highly sophisticated approach, Richards noted that the detection of them is much more challenging as a result. The infrastructure that supports these malicious attacks is typically hosted by providers who do not cooperate with law enforcement agencies, which complicates the process of dismantling these attacks.

Several European countries and the United States have increased sanctions in response to its growing threat, increasing the chance that these countries (including the United States) will use defenseless host hosting services, which are regularly exploited by cybercriminals to avoid legal action and avoid repercussions for their crimes. 

The American scientist Richards urges users to exercise extreme caution if they receive an email that appears to be coming from a legitimate organization and contains urgent requests for action that need to be taken immediately. As a precaution, users should not click on embedded links in emails, but instead should visit the official site to verify the authenticity of any alerts they receive or account-related issues. Taking a proactive approach effectively mitigates the threats posed by advanced phishing campaigns such as Astaroth.

Phishing URL Blocking Failure Leads to Cloudflare Service Disruptions

 


Yesterday, Cloudflare attempted to block an unintentional phishing URL within its R2 object storage platform, causing an outage that affected multiple services for nearly an hour. The outage was caused by an attempt to prevent spammers from accessing the URL. Its scalable and cost-efficient object storage service is comparable to Amazon's S3 and offers seamless integration into Cloudflare's ecosystem. 

As an S3-compatible storage service, the platform enables users to store their data across multiple locations, ensures data availability and reliability, and offers cost-free data retrievals, ensuring users can access their data without worries. A Cloudflare employee responded to an abuse report regarding a phishing URL hosted on its R2 platform, which caused the outage which occurred during the blackout. 

Inadvertently, the employee disabled the entire R2 Gateway service instead of restricting access to the specific endpoint, resulting in a significant service disruption. To prevent phishing URLs on the R2 platform, it accidentally resulted in a widespread outage of several Cloudflare services for almost an hour due to an attempt to block a phishing URL on the platform. 

Object storage solution Cloudflare R2 is no-egress-fee and has the same functionality as Amazon S3 and enables free data retrieval as well as S3 compatibility, replication, and seamless integration with other Cloudflare services to ensure efficiency and scalability in the storage of objects. In the incident which occurred late last week, Cloudflare employees responded to a complaint regarding a phishing URL hosted on the R2 platform.

However, the mitigation attempt resulted in an unintended disruption of service availability, which negatively impacted the operations of the platform. In the primary incident window of Cloudflare R2, all users were experiencing 100% failure rates when accessing their buckets and objects within the platform. Specifically, services that relied on R2 were experiencing higher error rates and operational failures as a result of their particular usage of the platform, as explained in the table below. 

Cloudflare R2 Object Storage and several related services were affected by an incident which took place from 08:10 to 09:09 UTC and lasted for 59 minutes. As a result of the impacted service failures, Stream experienced an entirely complete failure in video uploads and streaming, whereas Images experienced a 100% failure rate in uploads and downloads of images. During the week, Cache Reserve was completely down, raising origin requests to an all-time high. 

It has been observed that Vectorize experienced 75% failure rates for queries and failed to accomplish inserts, upserts, and deletes. It also experienced a 100% failure rate for insert, upsert, and delete operations. Log Delivery suffered delays and data loss, with up to 13.6% of all logs for R2-related jobs and up to 4.5% for non-R2 delivery jobs. Furthermore, the Key Transparency Auditor's signature publishing and reading operations were completely inoperable. Several other services were indirectly affected, experiencing partial disruptions, but they were not directly impacted. 

The error rates at Durable Objects increased by 0.9% following a service restoration due to reconnections, whereas Cache Purge experienced 1.8% more HTTP 5xx errors, as well as a tenfold increase in latency, as well as Workers & Pages experiencing a deployment failure rate of 0.002%, which was specifically affecting R2 projects only. As a consequence of the outage, all operations involving the R2 platform failed between 08:14 UTC and 09:13 UTC, meaning that 100% of operations involving R2 encountered errors. 

Services reliant on the R2 platform also saw an increase in the failure rate for operations that depend on it. During the period between 09:13 and 09:36 UTC, when R2 systems had recovered, and client connections had been restored, a backlog of requests caused a temporary increase in the operational load on the metadata layer of R2 based on Durable Objects. In North America, it was observed that there was only a 0.09% increase in error rates observed during this period, indicating that the impact was less severe. 

According to CloudFlare, the incident was primarily caused by human error and the absence of critical safeguards, such as validation checks for high-impact actions. The company has taken immediate corrective measures in response to the issue. These include removing the capability of disabling systems from the abuse review interface and limiting access to the Admin API so that internal accounts can no longer shut down services. 

Cloudflare's provisioning processes will be improved to reduce the risk of recurrence in the future, and stricter access controls will be enforced further to mitigate the risk of repeat incidents in the future. Additionally, two-party approval systems will be implemented for high-risk actions to further mitigate risk. The measures are intended to ensure the integrity of the system and prevent unintended interruptions of service as a result of these actions.

Quishing Scams Exploit Microsoft Sway Platform

 


It has been discovered that a new phishing campaign is being run using Microsoft Sway, which has been found by researchers. A series of attacks have been called the "Quishing" campaign to describe what is happening. The practice of "squishing" is a form of phishing that uses QR codes to lead people to malicious websites. An example of Quishing is embedding malicious URLs into a QR code to commit phishing. 

A few groups of victims in Asia and North America are primarily focusing on the campaign. In late December, researchers noticed that an unexpected spike in traffic to unique Microsoft Sway phishing pages arose as a result of a campaign called "quishing," which targeted Microsoft Office credentials.  As defined by Netskope Threat Labs, quishing is essentially phishing to trick users into opening malicious pages by presenting them with QR codes, which are commonly used in many forms of phishing. 

According to a spokesperson for the campaign, the campaign mainly targets victims in Asia and North America, across multiple industries such as the technology, manufacturing, and finance sectors. A researcher from the University of California, Davis, reported that "attackers instruct their victims to scan QR codes with their mobile devices, in the hope that these portable devices do not possess the strict security measures found on corporate-issued devices," according to an article written by the researchers. 

This QR phishing campaign utilizes two techniques that have been discussed in previous articles: transparent phishing in conjunction with Cloudflare Turnstile" Those who operate phishing websites use Cloudflare Turnstile to ensure that their malicious websites are protected from static analysis tools so that they can hide their malicious payloads, prevent web filtering providers from blocking their domains, and maintain a clean reputation among the web community. 

This is known as an attack-in-the-middle phishing technique, which is more sophisticated than traditional phishing techniques. The attackers not only attempt to gain access to the victims' credentials but also attempt to log them into the legitimate service using those credentials, bypassing multi-factor authentication, so they can steal sensitive tokens or cookies which can be used to gain further unauthorized access to the system. 

This is a massive QR code phishing campaign, which abused Microsoft Sway, a cloud-based tool for creating presentations online, to create landing pages that scammed Microsoft 365 users into handing over their credentials in exchange for money. According to Netskope Threat Labs, these attacks were spotted in July 2024 after detecting an increase of 2,000-fold in attacks exploiting Microsoft Sway to host phishing pages that allegedly steal access credentials for Microsoft 365 accounts. 

Interestingly, this surge of activity dates back to the first half of the year when minimal activity was reported. So, it comes as no surprise that this campaign has been so widespread. Essentially, they were interested in targeting users in Asia and North America, concentrating primarily on the technology, manufacturing, and finance sectors, which were the most likely to present themselves to them. A free application, called Sway, is available in Microsoft 365 for anyone with a Microsoft account who has a Microsoft account. 

Attackers, however, utilize this open access as an opportunity to fool users by misrepresenting them as legitimate cloud applications, thus defrauding them of the money they are paid to use them. Furthermore, Sway is accessed once an individual logs into their Microsoft 365 account, adding a layer of legitimacy to the attack, since it is accessible once the victim has already logged into the account, thus increasing the chances of them opening malicious links. 

Netskope Threat Labs identified a new QR code phishing campaign in July 2024, marking a significant development in cyber threats. This campaign primarily targets victims in Asia and North America, affecting various sectors, including manufacturing, technology, and finance. Cybercriminals employ diverse sharing methods, such as email, links, and social media platforms like Twitter, to direct users to phishing pages hosted on the sway. cloud.Microsoft domain. 

Once on these pages, victims are prompted to scan QR codes that subsequently lead them to malicious websites. Microsoft Sway, a platform known for its versatility, has been exploited in the past for phishing activities. Notably, five years ago, the PerSwaysion phishing campaign leveraged Microsoft Sway to target Office 365 login credentials. This campaign, driven by a phishing kit offered through a malware-as-a-service (MaaS) operation, was uncovered by Group-IB security researchers.

The attacks deceived at least 156 high-ranking individuals within small and medium-sized financial services companies, law firms, and real estate groups. The compromised accounts included those of executives, presidents, and managing directors across the U.S., Canada, Germany, the U.K., the Netherlands, Hong Kong, and Singapore. This escalation in phishing tactics highlights the ongoing battle between cybercriminals and cybersecurity professionals, where each defensive measure is met with a corresponding offensive innovation. 

The need for a comprehensive approach to cybersecurity has never been more apparent, as malicious actors continue to exploit seemingly innocuous technologies for nefarious purposes. With the rising popularity of Unicode QR code phishing techniques, security experts emphasize the importance of enhancing detection capabilities to analyze not just images but also text-based codes and other unconventional formats used to deceive users and infiltrate systems. This sophisticated phishing method underscores the continuous vigilance required to safeguard digital environments against increasingly cunning cyber threats.

Iranian Threat Actor TA453 Targets Jewish Figure with Fake Podcast Invite in Malicious Campaign

 

A recent cyber campaign by the Iranian threat actor TA453 has drawn significant attention following their targeting of a prominent Jewish religious figure with a fake podcast interview invitation. The campaign, which began in July 2024, involved a series of deceptive emails promoting a supposed podcast titled “Exploring Jewish Life in the Muslim World.” The attackers masqueraded as representatives of the Institute for the Study of War (ISW), a legitimate American non-profit think tank focused on military and foreign affairs research. 

On July 22, 2024, TA453 initiated contact with the target by sending an email from an address claiming to represent ISW’s Research Director. The email invited the recipient to participate in the podcast, a lure that successfully engaged the target. After initial correspondence, TA453 sent a DocSend URL containing a password-protected text file with a legitimate ISW podcast link. Researchers from Proofpoint believe this initial interaction was intended to build trust with the target, making them more likely to click on malicious links in future communications. 

Following the initial lure, TA453 escalated their attack by sending a Google Drive URL that led to a ZIP archive. This archive contained a malicious LNK file, which, when opened, deployed the BlackSmith toolset, including the AnvilEcho PowerShell trojan. AnvilEcho is a sophisticated malware capable of intelligence gathering and data exfiltration. It employs encryption and network communication techniques to evade detection, integrating multiple capabilities within a single PowerShell script. The trojan’s command-and-control (C2) infrastructure is hosted on a domain linked to previous TA453 operations. 

AnvilEcho continuously fetches and executes commands from the remote server via its “Do-It” function, which handles various tasks, including network connectivity, file manipulation, screenshot capture, and audio recording. The “Redo-It” function, located at the end of the malware’s code, orchestrates these commands while also collecting system reconnaissance data such as antivirus status, operating system details, and user information. According to researchers, the activities of TA453 are likely aimed at supporting intelligence collection for the Iranian government, specifically the Islamic Revolutionary Guard Corps’ Intelligence Organization. 

The tactics employed in this campaign bear a strong resemblance to those used by the Charming Kitten advanced persistent threat (APT) group, another Iranian cyber espionage unit. This operation is a classic example of multi-persona impersonation, where threat actors leverage legitimate links to build trust with victims before launching more harmful attacks.