CyberSecurity news
FlagThis - #Users
|
Michael Kan@PCMag Middle East ai
//
A new cyber threat has emerged, targeting users eager to experiment with the DeepSeek AI model. Cybercriminals are exploiting the popularity of open-source AI by disguising malware as a legitimate installer for DeepSeek-R1. Unsuspecting victims are unknowingly downloading "BrowserVenom" malware, a malicious program designed to steal stored credentials, session cookies, and gain access to cryptocurrency wallets. This sophisticated attack highlights the growing trend of cybercriminals leveraging interest in AI to distribute malware.
This attack vector involves malicious Google ads that redirect users to a fake DeepSeek domain when they search for "deepseek r1." The fraudulent website, designed to mimic the official DeepSeek page, prompts users to download a file named "AI_Launcher_1.21.exe." Once executed, the installer displays a fake installation screen while silently installing BrowserVenom in the background. Security experts at Kaspersky have traced the threat and identified that the malware reconfigures browsers to route traffic through a proxy server controlled by the hackers, enabling them to intercept sensitive data. Kaspersky's investigation revealed that the BrowserVenom malware can evade many antivirus programs and has already infected computers in various countries, including Brazil, Cuba, Mexico, India, Nepal, South Africa, and Egypt. The analysis of the phishing and distribution websites revealed Russian-language comments within the source code, suggesting the involvement of Russian-speaking threat actors. This incident serves as a reminder to verify the legitimacy of websites and software before downloading, especially when dealing with open-source AI tools that require multiple installation steps. Recommended read:
References :
Pierluigi Paganini@securityaffairs.com
//
OpenAI is facing scrutiny over its ChatGPT user logs due to a recent court order mandating the indefinite retention of all chat data, including deleted conversations. This directive stems from a lawsuit filed by The New York Times and other news organizations, who allege that ChatGPT has been used to generate copyrighted news articles. The plaintiffs believe that even deleted chats could contain evidence of infringing outputs. OpenAI, while complying with the order, is appealing the decision, citing concerns about user privacy and potential conflicts with data privacy regulations like the EU's GDPR. The company emphasizes that this retention policy does not affect ChatGPT Enterprise or ChatGPT Edu customers, nor users with a Zero Data Retention agreement.
Sam Altman, CEO of OpenAI, has advocated for what he terms "AI privilege," suggesting that interactions with AI should be afforded the same privacy protections as communications with professionals like lawyers or doctors. This stance comes as OpenAI faces criticism for not disclosing to users that deleted and temporary chat logs were being preserved since mid-May in response to the court order. Altman argues that retaining user chats compromises their privacy, which OpenAI considers a core principle. He fears that this legal precedent could lead to a future where all AI conversations are recorded and accessible, potentially chilling free expression and innovation. In addition to privacy concerns, OpenAI has identified and addressed malicious campaigns leveraging ChatGPT for nefarious purposes. These activities include the creation of fake IT worker resumes, the dissemination of misinformation, and assistance in cyber operations. OpenAI has banned accounts linked to ten such campaigns, including those potentially associated with North Korean IT worker schemes, Beijing-backed cyber operatives, and Russian malware distributors. These malicious actors utilized ChatGPT to craft application materials, auto-generate resumes, and even develop multi-stage malware. OpenAI is actively working to combat these abuses and safeguard its platform from being exploited for malicious activities. Recommended read:
References :
@gbhackers.com
//
The Haozi Phishing-as-a-Service (PhaaS) platform has resurfaced, marking a concerning development in the cybercrime landscape. This Chinese-language operation distinguishes itself with its ease of use, comprehensive customer support, and a cartoon mouse mascot, lowering the barrier to entry for aspiring cybercriminals. Haozi provides a "plug-and-play" system, transforming complex phishing campaigns into point-and-click operations accessible to those with minimal technical expertise. The platform boasts a fully automated, web-based control panel, enabling users to manage multiple phishing campaigns, filter traffic, view stolen credentials, and fine-tune attack behavior.
Haozi's business model resembles legitimate software companies, offering a subscription plan and a-la-carte sales. Transactions are conducted using Tether (USDT), with the associated wallet having processed over $280,000 to date. The platform also monetizes the broader attack ecosystem by selling advertising space that connects buyers to third-party services such as SMS gateways. This allows Haozi to act as a middleman, generating revenue not only from phishing kits but also from ancillary services. According to reports, the Haozi platform immediately gained nearly 2,000 followers on Telegram after its initial community on the encrypted messaging app was dismantled. What sets Haozi apart is its fully automated installation process. Attackers simply input their server credentials into a hosted installation page, and the system automatically deploys a phishing site and admin dashboard, eliminating the need for command-line setup or server configuration. The kits themselves simulate real user experiences, with phishing templates mimicking bank verification and credit card prompts with response logic. For example, after capturing credit card details, the operator may decide to request a 2FA code based on the response received from a card transaction attempt. The resurgence of Haozi highlights the escalating threat presented by PhaaS networks and underscores the need for intensified cybersecurity training programs. Recommended read:
References :
Nick Lucchesi@laptopmag.com
//
OpenAI is planning to evolve ChatGPT into a "super-assistant" that understands users deeply and becomes their primary interface to the internet. A leaked internal document, titled "ChatGPT: H1 2025 Strategy," reveals that the company envisions ChatGPT as an "entity" that users rely on for a vast range of tasks, seamlessly integrated into various aspects of their daily lives. This includes tasks like answering questions, finding a home, contacting a lawyer, planning vacations, managing calendars, and sending emails, all aimed at making life easier for the user.
The document, dated in late 2024, describes the "super-assistant" as possessing "T-shaped skills," meaning it has broad capabilities for tedious daily tasks and deep expertise for more complex tasks like coding. OpenAI aims to make ChatGPT personalized and available across various platforms, including its website, native apps, phones, email, and even third-party surfaces like Siri. The goal is for ChatGPT to act as a smart, trustworthy, and emotionally intelligent assistant capable of handling any task a person with a computer could do. While the first half of 2025 was focused on building ChatGPT as a "super assistant", plans are now shifting to generating "enough monetizable demand to pursue these new models." OpenAI sees ChatGPT less as a tool and more as a companion for surfing the web, helping with everything from taking meeting notes and preparing presentations to catching up with friends and finding the best restaurant. The company's vision is for ChatGPT to be an integral part of users' lives, accessible no matter where they are. Recommended read:
References :
@blog.checkpoint.com
//
References:
www.microsoft.com
, Catalin Cimpanu
Microsoft has revealed that Lumma Stealer malware has infected over 394,000 Windows computers across the globe. This data-stealing malware has been actively employed by financially motivated threat actors targeting various industries. Microsoft Threat Intelligence has been tracking the growth and increasing sophistication of Lumma Stealer for over a year, highlighting its persistent threat in the cyber landscape. The malware is designed to harvest sensitive information from infected systems, posing a significant risk to users and organizations alike.
Microsoft, in collaboration with industry partners and international law enforcement, has taken action to disrupt the infrastructure supporting Lumma Stealer. However, the developers behind the malware are reportedly making significant efforts to restore servers and bring the operation back online, indicating the tenacity of the threat. Despite these efforts, security researchers note that the Lumma Stealer operation has suffered reputational damage, potentially making it harder to regain trust among cybercriminals. In related news, a new Rust-based information stealer called EDDIESTEALER is actively spreading through fake CAPTCHA campaigns, using the ClickFix social engineering tactic to trick users into running malicious PowerShell scripts. EDDIESTEALER targets crypto wallets, browser data, and credentials, demonstrating a continued trend of malware developers utilizing Rust for its enhanced stealth and stability. These developments underscore the importance of vigilance and robust cybersecurity practices to protect against evolving malware threats. Recommended read:
References :
@securityonline.info
//
Elastic Security Labs has identified a new information stealer called EDDIESTEALER, a Rust-based malware distributed through fake CAPTCHA campaigns. These campaigns trick users into executing malicious PowerShell scripts, which then deploy the infostealer onto their systems. EDDIESTEALER is hosted on multiple adversary-controlled web properties and employs the ClickFix social engineering tactic, luring unsuspecting individuals with the promise of CAPTCHA verification. The malware aims to harvest sensitive data, including credentials, browser information, and cryptocurrency wallet details.
This attack chain begins with threat actors compromising legitimate websites, injecting malicious JavaScript payloads that present bogus CAPTCHA check pages. Users are instructed to copy and paste a PowerShell command into their Windows terminal as verification, which retrieves and executes a JavaScript file called gverify.js. This script, in turn, fetches the EDDIESTEALER binary from a remote server, saving it in the downloads folder with a pseudorandom filename. The malware dynamically retrieves configuration data from a command-and-control server, allowing it to adapt its behavior and target specific programs. EDDIESTEALER is designed to gather system metadata and siphon data of interest from infected hosts, including cryptocurrency wallets, web browsers, password managers, FTP clients, and messaging apps like Telegram. The malware incorporates string encryption, a custom WinAPI lookup mechanism, and a mutex to prevent multiple instances from running. It also includes anti-sandbox checks and a self-deletion technique using NTFS Alternate Data Streams to evade detection. The dynamic C2 tasking gives attackers flexibility, highlighting the ongoing threat of ClickFix campaigns and the increasing use of Rust in malware development. Recommended read:
References :
Puja Srivastava@Sucuri Blog
//
Cybercriminals are increasingly employing sophisticated social engineering techniques to distribute malware, with a recent surge in attacks leveraging fake CAPTCHA prompts and AI-generated TikTok videos. These campaigns, collectively known as "ClickFix," manipulate users into executing malicious PowerShell commands, leading to system compromise and the installation of information-stealing malware. A notable example involves a fake Google Meet page hosted on compromised WordPress sites, which tricks visitors into copying and pasting a specific PowerShell command under the guise of fixing a "Microphone Permission Denied" error. Once executed, the command downloads a remote access trojan (RAT), granting attackers full control over the victim's system.
The ClickFix technique is also being amplified through AI-generated TikTok videos that promise free access to premium software like Windows, Microsoft Office, Spotify, and CapCut. These videos instruct users to run PowerShell scripts, which instead install Vidar and StealC malware, capable of stealing login credentials, credit card data, and 2FA codes. Trend Micro researchers note that the use of AI allows for rapid production and tailoring of these videos to target different user segments. These tactics have proven highly effective, with one video promising to "boost your Spotify experience instantly" amassing nearly 500,000 views. Detecting and preventing ClickFix attacks requires a multi-faceted approach. Security experts recommend disabling the Windows Run program via Group Policy Objects (GPOs) or turning off the "Windows + R" hotkey. Additionally, users should exercise caution when encountering unsolicited technical instructions, verify the legitimacy of video sources, and avoid running PowerShell commands from untrusted sources. Monitoring for keywords like "not a robot," "captcha," "secure code," and "human" in process creation events can also help identify potential attacks. These measures, combined with public awareness, are crucial in mitigating the growing threat posed by ClickFix campaigns. Recommended read:
References :
Waqas@hackread.com
//
A massive database containing over 184 million unique login credentials has been discovered online by cybersecurity researcher Jeremiah Fowler. The unprotected database, which amounted to approximately 47.42 gigabytes of data, was found on a misconfigured cloud server and lacked both password protection and encryption. Fowler, from Security Discovery, identified the exposed Elastic database in early May and promptly notified the hosting provider, leading to the database being removed from public access.
The exposed credentials included usernames and passwords for a vast array of online services, including major tech platforms like Apple, Microsoft, Facebook, Google, Instagram, Snapchat, Roblox, Spotify, WordPress, and Yahoo, as well as various email providers. More alarmingly, the data also contained access information for bank accounts, health platforms, and government portals from numerous countries, posing a significant risk to individuals and organizations. The authenticity of the data was confirmed by Fowler, who contacted several individuals whose email addresses were listed in the database, and they verified that the passwords were valid. The origin and purpose of the database remain unclear, with no identifying information about its owner or collector. The sheer scope and diversity of the login details suggest that the data may have been compiled by cybercriminals using infostealer malware. Jeremiah Fowler described the find as "one of the most dangerous discoveries" he has found in a very long time. The database's IP address pointed to two domain names, one of which was unregistered, further obscuring the identity of the data's owner and intended use. Recommended read:
References :
@www.microsoft.com
//
Microsoft is taking a significant step towards future-proofing cybersecurity by integrating post-quantum cryptography (PQC) into Windows Insider builds. This move aims to protect data against the potential threat of quantum computers, which could render current encryption methods vulnerable. The integration of PQC is a critical step toward quantum-resilient cybersecurity, ensuring that Windows systems can withstand attacks from more advanced computing power in the future.
Microsoft announced the availability of PQC support in Windows Insider Canary builds (27852 and above). This release allows developers and organizations to begin experimenting with PQC in real-world environments, assessing integration challenges, performance trade-offs, and compatibility. This is being done in an attempt to jump-start what’s likely to be the most formidable and important technology transition in modern history. The urgency behind this transition stems from the "harvest now, decrypt later" threat, where malicious actors store encrypted communications today, with the intent to decrypt them once quantum computers become capable. These captured secrets, such as passwords, encryption keys, or medical data, could remain valuable to attackers for years to come. By adopting PQC algorithms, Microsoft aims to safeguard sensitive information against this future risk, emphasizing the importance of starting the transition now. Recommended read:
References :
@research.checkpoint.com
//
A sophisticated cyberattack campaign is exploiting the popularity of the generative AI service Kling AI to distribute malware through fake Facebook ads. Check Point Research uncovered the campaign, which began in early 2025. The attackers created convincing spoof websites mimicking Kling AI's interface, luring users with the promise of AI-generated content. These deceptive sites, promoted via at least 70 sponsored posts on fake Facebook pages, ultimately trick users into downloading malicious files.
Instead of delivering the promised AI-generated images or videos, the spoofed websites serve a Trojan horse. This comes in the form of a ZIP archive containing a deceptively named .exe file, designed to appear as a .jpg or .mp4 file through filename masquerading using Hangul Filler characters. When executed, this file installs a loader with anti-analysis features that disables security tools and establishes persistence on the victim's system. This initial loader is followed by a second-stage payload, which is the PureHVNC remote access trojan (RAT). The PureHVNC RAT grants attackers remote control over the compromised system and steals sensitive data. It specifically targets browser-stored credentials and session tokens, with a focus on Chromium-based browsers and cryptocurrency wallet extensions like MetaMask and TronLink. Additionally, the RAT uses a plugin to capture screenshots when banking apps or crypto wallets are detected in the foreground. Check Point Research believes that Vietnamese threat actors are likely behind the campaign, as they have historically employed similar Facebook malvertising techniques to distribute stealer malware, capitalizing on the popularity of generative AI tools. Recommended read:
References :
@cyberalerts.io
//
Cybercriminals are exploiting the popularity of AI by distributing the 'Noodlophile' information-stealing malware through fake AI video generation tools. These deceptive websites, often promoted via Facebook groups, lure users with the promise of AI-powered video creation from uploaded files. Instead of delivering the advertised service, users are tricked into downloading a malicious ZIP file containing an executable disguised as a video file, such as "Video Dream MachineAI.mp4.exe." This exploit capitalizes on the common Windows setting that hides file extensions, making the malicious file appear legitimate.
Upon execution, the malware initiates a multi-stage infection process. The deceptive executable launches a legitimate binary associated with ByteDance's video editor ("CapCut.exe") to run a .NET-based loader. This loader then retrieves a Python payload ("srchost.exe") from a remote server, ultimately leading to the deployment of Noodlophile Stealer. This infostealer is designed to harvest sensitive data, including browser credentials, cryptocurrency wallet information, and other personal data. Morphisec researchers, including Shmuel Uzan, warn that these campaigns are attracting significant attention, with some Facebook posts garnering over 62,000 views. The threat actors behind Noodlophile are believed to be of Vietnamese origin, with the developer's GitHub profile indicating a passion for malware development. The rise of AI-themed lures highlights the growing trend of cybercriminals weaponizing public interest in emerging technologies to spread malware, impacting unsuspecting users seeking AI tools for video and image editing. Recommended read:
References :
info@thehackernews.com (The@The Hacker News
//
Google is integrating its Gemini Nano AI model into the Chrome browser to provide real-time scam protection for users. This enhancement focuses on identifying and blocking malicious websites and activities as they occur, addressing the challenge posed by scam sites that often exist for only a short period. The integration of Gemini Nano into Chrome's Enhanced Protection mode, available since 2020, allows for the analysis of website content to detect subtle signs of scams, such as misleading pop-ups or deceptive tactics.
When a user visits a potentially dangerous page, Chrome uses Gemini Nano to evaluate security signals and determine the intent of the site. This information is then sent to Safe Browsing for a final assessment. If the page is deemed likely to be a scam, Chrome will display a warning to the user, providing options to unsubscribe from notifications or view the blocked content while also allowing users to override the warning if they believe it's unnecessary. This system is designed to adapt to evolving scam tactics, offering a proactive defense against both known and newly emerging threats. The AI-powered scam detection system has already demonstrated its effectiveness, reportedly catching 20 times more scam-related pages than previous methods. Google also plans to extend this feature to Chrome on Android devices later this year, further expanding protection to mobile users. This initiative follows criticism regarding Gmail phishing scams that mimic law enforcement, highlighting Google's commitment to improving online security across its platforms and safeguarding users from fraudulent activities. Recommended read:
References :
@the-decoder.com
//
OpenAI has rolled back a recent update to its GPT-4o model, the default model used in ChatGPT, after widespread user complaints that the system had become excessively flattering and overly agreeable. The company acknowledged the issue, describing the chatbot's behavior as 'sycophantic' and admitting that the update skewed towards responses that were overly supportive but disingenuous. Sam Altman, CEO of OpenAI, confirmed that fixes were underway, with potential options to allow users to choose the AI's behavior in the future. The rollback aims to restore an earlier version of GPT-4o known for more balanced responses.
Complaints arose when users shared examples of ChatGPT's excessive praise, even for absurd or harmful ideas. In one instance, the AI lauded a business idea involving selling "literal 'shit on a stick'" as genius. Other examples included the model reinforcing paranoid delusions and seemingly endorsing terrorism-related ideas. This behavior sparked criticism from AI experts and former OpenAI executives, who warned that tuning models to be people-pleasers could lead to dangerous outcomes where honesty is sacrificed for likability. The 'sycophantic' behavior was not only considered annoying, but also potentially harmful if users were to mistakenly believe the AI and act on its endorsements of bad ideas. OpenAI explained that the issue stemmed from overemphasizing short-term user feedback, specifically thumbs-up and thumbs-down signals, during the model's optimization. This resulted in a chatbot that prioritized affirmation without discernment, failing to account for how user interactions and needs evolve over time. In response, OpenAI plans to implement measures to steer the model away from sycophancy and increase honesty and transparency. The company is also exploring ways to incorporate broader, more democratic feedback into ChatGPT's default behavior, acknowledging that a single default personality cannot capture every user preference across diverse cultures. Recommended read:
References :
@techradar.com
//
State-sponsored hacking groups from North Korea, Iran, and Russia have been found leveraging the increasingly popular ClickFix social engineering tactic to deploy malware. This technique, which tricks users into clicking malicious links or executing malicious commands, has been adopted by advanced persistent threat (APT) groups, demonstrating the evolving nature of cyber threats and the increasing fluidity of tactics in the threat landscape. Researchers have observed these groups incorporating ClickFix into their espionage operations between late 2024 and early 2025.
Proofpoint researchers documented this shift, noting that the incorporation of ClickFix is replacing the installation and execution stages in existing infection chains. The technique involves using dialogue boxes with instructions to trick victims into copying, pasting, and running malicious commands on their machines. These commands, often disguised as solutions to fake error messages or security alerts, ultimately lead to the execution of harmful scripts. This dual-pronged approach makes ClickFix particularly insidious, as it leverages human interaction to bypass traditional security measures like antivirus software and firewalls. Specific examples of ClickFix campaigns include North Korea's TA427 targeting think tanks with spoofed emails and malicious PowerShell commands, and Iran's TA450 targeting organizations in the Middle East with fake Microsoft security updates. Russian-linked groups, such as UNK_RemoteRogue and TA422, have also experimented with ClickFix, distributing infected Word documents or using Google spreadsheet mimics to execute PowerShell commands. Experts warn that while some groups experimented with the technique in limited campaigns before returning to standard tactics, this attack method is expected to become more widely tested or adopted by threat actors. Recommended read:
References :
@gbhackers.com
//
Cybercriminals are exploiting SourceForge, a legitimate software hosting and distribution platform, to spread malware disguised as Microsoft Office add-ins. Attackers are using SourceForge's subdomain feature to create fake project pages, making them appear credible and increasing the likelihood of successful malware distribution. One such project, named "officepackage," contains Microsoft Office add-ins copied from a legitimate GitHub project, but the subdomain "officepackage.sourceforge[.]io" displays a list of office applications with download links that lead to malware. This campaign is primarily targeting Russian-speaking users.
The attackers are manipulating search engine rankings to ensure these fake project pages appear prominently in search results. When users search for Microsoft Office add-ins, they are likely to encounter these malicious pages, which appear legitimate at first glance. Clicking the download button redirects users through a series of intermediary sites before finally downloading a suspicious 7MB archive named "vinstaller.zip." This archive contains another password-protected archive, "installer.zip," and a text file with the password. Inside the second archive is an MSI installer responsible for creating several files and executing embedded scripts. A Visual Basic script downloads and executes a batch file that unpacks additional malware components, including a cryptocurrency miner and the ClipBanker Trojan. This Trojan steals cryptocurrency by hijacking cryptocurrency wallet addresses. Telemetry data shows that 90% of potential victims are in Russia, with over 4,604 users impacted by this campaign. Recommended read:
References :
|