This is a monthly newsletter that I share internally with our Cybersecurity team. Feel free to take and use for your own team. Created with the help of ChatGPT.

ModeLeak Vulnerabilities in Google's Vertex AI Platform 

Palo Alto Networks' Unit 42 team has uncovered two critical vulnerabilities, collectively termed "ModeLeak," within Google's Vertex AI platform. These flaws could enable attackers to escalate privileges and exfiltrate sensitive machine learning (ML) models, including fine-tuned large language model (LLM) adapters. 

Key Insights: 

• Privilege Escalation via Custom Jobs: Attackers can exploit custom job permissions to gain unauthorized access to data services within a project, leading to potential exposure of sensitive information. 

• Model Exfiltration through Malicious Models: By deploying a poisoned model, adversaries can exfiltrate other fine-tuned models in the environment, risking proprietary data and custom optimizations. 

Google has addressed these vulnerabilities by implementing fixes in the Vertex AI platform. Organizations utilizing Vertex AI should review their security protocols to ensure protection against similar threats. 

Further Reading: Unit 42 Blog 

 

 

Black Basta Ransomware Adopts Advanced Social Engineering Tactics 

The Black Basta ransomware group has recently enhanced its attack strategies by incorporating sophisticated social engineering techniques, including email bombing, QR code phishing, and the deployment of custom malware payloads. 

Key Developments: 

• Email Bombing: Attackers inundate targets with excessive emails by subscribing their addresses to numerous mailing lists. This tactic overwhelms victims and increases the likelihood of interaction with subsequent malicious communications. 

• Impersonation via Microsoft Teams: Threat actors pose as IT support personnel, contacting victims through Microsoft Teams to establish trust and facilitate the installation of remote access tools. 

• QR Code Phishing: Malicious QR codes are sent to victims, directing them to phishing sites designed to harvest credentials or deploy additional malware. 

• Custom Malware Deployment: The group utilizes bespoke tools such as KNOTWRAP (a memory-only dropper) and KNOTROCK (a .NET-based utility) to execute ransomware payloads stealthily. 

Further Reading: The Hacker News 

 

 

North Korean IT Workers Infiltrating Global Companies 

Recent investigations have uncovered that operatives from the Democratic People's Republic of Korea (DPRK) are securing remote IT positions in international companies under false identities. These individuals channel their earnings to fund North Korea's weapons programs, posing significant security and compliance risks to employers. 

Key Insights: 

• Use of False Identities: North Korean IT workers often utilize stolen or fabricated identities to obtain employment, making detection challenging. 

• Revenue Generation for DPRK: Earnings from these positions are funneled back to North Korea, supporting its sanctioned weapons development initiatives. 

• Potential for Insider Threats: Beyond financial implications, these operatives may have access to sensitive company data, increasing the risk of intellectual property theft and cyber espionage. 

Further Reading: Unit 42 Blog 

 

 

North Korean IT Workers Linked to Phishing Attacks via Malicious Video Conferencing Apps 

Unit 42 researchers have identified a cluster of North Korean IT operatives, designated as CL-STA-0237, involved in phishing attacks that deploy malware through counterfeit video conferencing applications. Operating primarily from Laos, these individuals have secured positions in various companies, leveraging their roles to further malicious activities. 

Key Insights: 

• Malware Distribution: The group utilizes fraudulent video conferencing platforms to disseminate malware, notably the BeaverTail and InvisibleFerret remote access trojans, compromising systems during supposed job interview processes. 

• Global Reach: By infiltrating organizations worldwide, these operatives support North Korea's illicit endeavors, including its weapons of mass destruction and ballistic missile programs. 

• Evolving Tactics: The shift from merely seeking income to engaging in aggressive malware campaigns indicates a significant escalation in their operational strategies. 

Further Reading: Unit 42 Blog 

 

 

Surge in 'ClickFix' Social Engineering Attacks 

Cybersecurity researchers have identified a significant increase in the use of a social engineering tactic known as "ClickFix." This method deceives users into copying and pasting malicious commands into their systems, leading to malware infections. 

Key Developments: 

• Deceptive Error Messages: Attackers present fake error dialogs, prompting users to execute provided commands to resolve non-existent issues. 

• Malware Delivery: By following these instructions, users inadvertently run scripts that download and install malware such as Lumma Stealer and AsyncRAT. 

• Global Impact: Campaigns employing ClickFix techniques have targeted organizations worldwide, with notable incidents involving fake GitHub security notifications and counterfeit software updates. 

Further Reading: Proofpoint Blog 

 

 

Malicious Ads Deliver SocGholish Malware to Kaiser Permanente Employees 

A recent cyberattack has targeted Kaiser Permanente employees through malicious advertisements on Google Search, leading to the distribution of SocGholish malware. 

Key Developments: 

• Malicious Advertisements: Threat actors placed deceptive ads mimicking Kaiser Permanente's HR portal to lure employees searching for benefits and payroll information. 

• Compromised Website Redirects: Clicking the fraudulent ad redirected users to a compromised website, bellonasoftware[.]com, which briefly displayed a phishing page before prompting a fake browser update. 

• SocGholish Malware Deployment: The fake browser update led to the download of "Update.js," a malicious script associated with the SocGholish malware campaign, designed to collect system information and potentially allow human operators to execute further malicious actions. 

This incident highlights the evolving tactics of cybercriminals in exploiting trusted platforms like Google Ads to distribute malware. 

Further Reading: Malwarebytes Blog 

 

 

DarkGate Malware Leveraging Vishing via Microsoft Teams 

Recent analyses have identified a concerning trend in which cybercriminals are deploying DarkGate malware through vishing (voice phishing) attacks conducted via Microsoft Teams. 

Key Developments: 

• Social Engineering Tactics: Attackers impersonate employees from known client organizations during Microsoft Teams calls, convincing victims to download remote desktop applications like AnyDesk. 

• Malware Deployment: Once remote access is established, DarkGate malware is installed, enabling threat actors to execute malicious commands, gather system information, and maintain persistent access. 

• Operational Impact: Although some attacks have been thwarted before data exfiltration, the initial breach underscores vulnerabilities in user awareness and the potential for significant security incidents. 

Further Reading: Trend Micro Research 

 

 

 

Sophisticated Phishing Campaigns Exploit Trusted Platforms 

Recent analyses have uncovered advanced phishing campaigns targeting employees across multiple industries and jurisdictions. These operations employ sophisticated techniques to bypass Secure Email Gateways (SEGs) and exploit trusted platforms, creating highly convincing schemes to deceive victims and steal their credentials. 

Key Developments: 

• Exploitation of Trusted Platforms: Attackers leverage familiar platforms and services to enhance the credibility of their phishing attempts, making it more challenging for victims to identify fraudulent communications. 

• Bypassing Secure Email Gateways (SEGs): The campaigns utilize advanced methods to evade detection by SEGs, allowing malicious emails to reach employees' inboxes undetected. 

• Wide-Ranging Targets: Over 30 companies across 12 industries and 15 jurisdictions have been affected, indicating a broad and indiscriminate approach by the threat actors. 

Further Reading: Group-IB Blog 

 

 

Top Cyber Attacker Techniques (August–October 2024) 

Recent analyses have identified key cyber attacker tactics, techniques, and procedures (TTPs) observed between August 1 and October 31, 2024. 

Key Developments: 

• Phishing Incidents: Phishing accounted for 46% of all customer incidents during this period, indicating a significant rise likely due to high employee turnover and the accessibility of phishing kits. 

• Prevalent Malware: "SocGholish" and "LummaC2" emerged as the most frequently observed malware in customer environments, highlighting their widespread use in recent attacks. 

• Cloud Services Alerts: There was a 20% increase in cloud services alerts, correlating with the rising adoption of cloud accounts and associated security challenges. 

• Ransomware Activity: Despite a slowdown in "LockBit" ransomware activity due to law enforcement actions and a loss of affiliate trust, it remains a key player. Meanwhile, "RansomHub" is rising rapidly due to its attractive ransomware-as-a-service (RaaS) model. The U.S., manufacturing sector, and professional, scientific, and technical services (PSTS) sector are primary targets amidst an overall increase in ransomware attacks. 

• Initial Access Broker (IAB) Activity: IAB activity increased by 16%, heavily targeting U.S.-based organizations, possibly due to perceived financial capabilities stemming from cyber insurance. 

• Insider Threat Content: A 7% rise in insider threat discussions on cybercrime forums was noted, driven by significant financial incentives, underscoring the growing complexity of cybersecurity challenges. 

• Impersonating Domain Alerts: There was a 6% increase in alerts related to impersonating domains, indicating ongoing reliance on simple techniques to capture credentials and data. 

Further Reading: ReliaQuest Blog 

 

 

Phishing Attacks Double in 2024 

Recent analyses reveal a significant surge in phishing activities throughout 2024, with overall phishing messages increasing by 202% in the latter half of the year. Notably, credential phishing attacks have escalated by 703% during the same period. 

Key Developments: 

• Prevalence of Zero-Day URLs: Approximately 80% of malicious links identified are zero-day threats—newly created URLs designed to evade traditional detection methods. 

• Diversification of Attack Vectors: While link-based phishing remains predominant, there is a notable increase in text-based threats, such as business email compromise (BEC) and invoice scams, as well as file-based threats employing techniques like HTML smuggling. 

• Expansion Beyond Email: Phishing attacks are increasingly targeting multiple platforms, including SMS, LinkedIn, and Microsoft Teams, indicating a shift towards multichannel approaches. 

Further Reading: Infosecurity Magazine 

 

 

Surge in Phishing Attacks via New Top-Level Domains 

Recent analyses reveal a significant increase in phishing attacks, with a 40% rise observed in the year ending August 2024. A substantial portion of this growth is attributed to the exploitation of new generic top-level domains (gTLDs) such as .shop, .top, and .xyz, which are favored by cybercriminals due to their low registration costs and minimal verification requirements. 

Key Developments: 

• Disproportionate Use in Cybercrime: Although new gTLDs constitute only 11% of the market for new domains, they account for approximately 37% of reported cybercrime domains between September 2023 and August 2024. 

• Attraction to Low-Cost Registrations: Registrars offering domain registrations for less than $1, with little to no identity verification, are particularly appealing to spammers and scammers seeking to conduct malicious activities anonymously. 

• ICANN's Expansion Plans: Despite the misuse of these new gTLDs, the Internet Corporation for Assigned Names and Numbers (ICANN) is proceeding with plans to introduce additional gTLDs, potentially broadening the landscape for cybercriminal activities. 

Further Reading: Krebs on Security 

 

 

Surge in Suspicious Domain Registrations Exploiting High-Profile Events 

Recent analyses have identified a significant increase in suspicious domain registration campaigns exploiting high-profile events, such as the 2024 Summer Olympics in Paris. 

Key Developments: 

• Event-Driven Domain Registrations: Threat actors register deceptive domains containing event-specific keywords to mimic official websites, aiming to deceive users seeking legitimate information. 

• Exploitation of Public Interest: Cybercriminals leverage global events to attract large audiences, using fraudulent domains to distribute malware, conduct phishing attacks, or sell counterfeit merchandise. 

• Indicators of Malicious Activity: Monitoring domain registrations, DNS traffic, URL patterns, and textual characteristics can help identify and mitigate these threats. 

Further Reading: Unit 42 Blog 

 

 

Zloader Malware Adopts DNS Tunneling for Stealthier C2 Communications 

Recent analyses have identified that the Zloader malware, a modular Trojan based on the leaked Zeus source code, has incorporated DNS tunneling into its command-and-control (C2) communication methods. 

Key Developments: 

• DNS Tunneling Implementation: Zloader now employs a custom protocol over DNS, utilizing IPv4 to tunnel encrypted TLS network traffic. This technique enables the malware to conceal its C2 communications within standard DNS queries and responses, making detection more challenging. 

• Enhanced Anti-Analysis Features: The latest version of Zloader includes improved anti-analysis capabilities, such as environment checks and API import resolution algorithms, to evade malware sandboxes and static detection methods. 

• Interactive Shell Capability: Zloader has introduced an interactive shell that supports over a dozen commands, potentially facilitating hands-on keyboard activity by threat actors during attacks. 

Further Reading: Zscaler Blog 

 

 

Cybercriminals Exploit Fake CAPTCHAs to Distribute Malware 

Recent analyses have identified a deceptive tactic where cybercriminals use fake CAPTCHA pages to distribute malware, exploiting users' trust in these verification systems. 

Key Developments: 

• Malicious Redirects: Users visiting compromised websites are redirected to fraudulent CAPTCHA pages that closely mimic legitimate services like Google and CloudFlare. 

• Clipboard Hijacking: These fake CAPTCHAs silently copy malicious commands to the user's clipboard via JavaScript, prompting them to execute these commands unknowingly through the Windows Run prompt. 

• Malware Installation: Executing the copied commands leads to the installation of malware, including information stealers and remote-access trojans (RATs), which can extract sensitive data and provide persistent access to compromised systems. 

Further Reading: ReliaQuest Blog 

 

 

Threat Actors Exploit LDAP for Network Enumeration 

Recent analyses have identified that both nation-state and cybercriminal threat actors are leveraging the Lightweight Directory Access Protocol (LDAP) to perform network enumeration within Active Directory environments. 

Key Developments: 

• Abuse of LDAP Attributes: Attackers utilize LDAP queries to extract sensitive information, such as user accounts, group memberships, and permissions, facilitating lateral movement and privilege escalation within compromised networks. 

• Use of Enumeration Tools: Tools like BloodHound and its data collector, SharpHound, are commonly employed to map Active Directory structures, identifying potential attack paths and high-value targets. 

• Detection Challenges: Distinguishing between legitimate and malicious LDAP activity is difficult due to the high volume of benign LDAP traffic in typical network environments, complicating efforts to detect and mitigate these attacks. 

Further Reading: Unit 42 Blog 

 

 

'Araneida' Web Hacking Service Linked to Turkish IT Firm 

Recent investigations have uncovered that 'Araneida,' a cloud-based web hacking service, is utilizing a cracked version of Acunetix—a commercial web application vulnerability scanner—to facilitate cyberattacks. Notably, this service has been traced back to a Turkish information technology firm. 

Key Developments: 

• Exploitation of Cracked Software: Araneida employs an unauthorized version of Acunetix, enabling users to perform offensive reconnaissance, extract user data, and identify exploitable vulnerabilities on target websites. 

• Proxy Integration for Anonymity: The service incorporates a robust proxy network, allowing scans to originate from a diverse pool of IP addresses, thereby concealing the true source of the activity. 

• Cybercriminal Promotion: Advertised on multiple cybercrime forums and boasting a Telegram channel with nearly 500 subscribers, Araneida has been linked to the compromise of over 30,000 websites within six months. One user claimed to have purchased a luxury vehicle using proceeds from payment card data obtained through the service. 

• Connection to Turkish IT Firm: Investigations reveal that the domain araneida[.]co, operational since February 2023, is associated with an individual employed as a senior software developer at Bilitro Yazilim, an IT firm based in Ankara, Turkey. 

Further Reading: Krebs on Security 

 

 

LLMs Employed to Obfuscate Malicious JavaScript 

Recent analyses have revealed that adversaries are leveraging large language models (LLMs) to obfuscate malicious JavaScript code, enhancing its ability to evade detection mechanisms. 

Key Developments: 

Automated Code Obfuscation: Attackers utilize LLMs to iteratively transform malicious JavaScript through techniques such as variable renaming, dead code insertion, and whitespace removal, without altering the code's functionality. 

Evasion of Detection Tools: These LLM-generated variants can bypass traditional detection tools, including static analysis models, by producing natural-looking code that is harder to identify as malicious. 

Scalability of Attacks: The use of LLMs enables the creation of numerous unique malware variants at scale, increasing the difficulty for security systems to detect and mitigate these threats effectively. 

Further Reading: Unit 42 Blog 

 

 

Mobile Phishing Attacks Employ New Tactics to Evade Security Measures 

Recent analyses have identified a novel social engineering tactic targeting mobile banking users. Attackers are leveraging Progressive Web Apps (PWAs) and WebAPKs to distribute phishing websites disguised as legitimate applications, effectively bypassing traditional security warnings and app store vetting processes. 

Key Insights: 

• Exploitation of PWAs and WebAPKs: Unlike traditional apps, these malicious PWAs and WebAPKs are essentially phishing websites packaged to look like legitimate applications. This means they do not exhibit the typical behaviors or characteristics associated with malware, making detection more challenging. 

• Bypassing Security Measures: Their ability to bypass traditional security warnings of a mobile operating system, and total sidestepping of app store vetting processes, is particularly concerning. This allows attackers to distribute malicious content without triggering standard security alerts. 

• Anticipated Increase in Sophistication: It is anticipated that more sophisticated and varied phishing campaigns utilizing PWAs and WebAPKs will emerge, unless mobile platforms change their approach towards them. 

Further Reading: KnowBe4 Blog