Cybersecurity Brief – 2025-07-18

Fri, 18 Jul 2025 00:00:00 +0000

Major Incidents or Breaches

Cryptocurrency exchange BigONE suffered a security breach resulting in the theft of digital assets valued at $27 million.
An adoption agency leaked over 1.1 million records, including names and sensitive information of children, birth parents, and adoptive parents.
Chinese state-sponsored group Salt Typhoon breached a U.S. Army National Guard network for nine months in 2024, exfiltrating network configuration files and credentials.
Security researchers revealed that poor password practices at Paradox.ai led to the exposure of personal information belonging to millions of job applicants.
A vulnerability in Meta AI (now fixed) could have allowed unauthorized access to private chatbot conversations.

Cisco disclosed and patched a critical pre-authentication vulnerability (CVE-2025-20337, CVSS 10) in Identity Services Engine (ISE) and ISE-PIC, allowing unauthenticated remote code execution and file storage.
Citrix NetScaler vulnerability “CitrixBleed 2” (CVE-2025-5777) was actively exploited weeks before public proof-of-concept exploits were released.
Oracle’s July 2025 Critical Patch Update addressed approximately 200 unique CVEs, with 309 security patches issued.
Google released a Chrome update patching six vulnerabilities, including one zero-day.
VMware patched four zero-day vulnerabilities in ESXi, Workstation, Fusion, and Tools, which were exploited at Pwn2Own Berlin 2025.
A known Apache HTTP Server vulnerability is being exploited to deploy the Linuxsys cryptocurrency miner.

Kaspersky identified “GhostContainer,” a new backdoor targeting Microsoft Exchange servers of high-value organizations in Asia, based on open-source tools.
Three previously undocumented Chinese state-sponsored threat actors, as well as a fourth group, targeted Taiwan’s semiconductor sector with spear-phishing, Cobalt Strike, and custom backdoors.
Threat actors are distributing Amadey malware and data stealers via malicious payloads hosted on public GitHub repositories to bypass security filters.
The Matanbuchus malware loader is being delivered through social engineering over Microsoft Teams voice calls, with attackers impersonating IT helpdesk staff.
LameHug malware uses large language models (LLMs) to generate Windows data-theft commands in real time.
Europol coordinated an international operation disrupting the infrastructure of the pro-Russian hacktivist group NoName057(16), known for DDoS attacks against Ukraine.
Research highlights increased use of DNS records to hide malicious code and conduct prompt injection attacks against chatbots.

PsMapExec, a new PowerShell-based tool, has been released for enumeration and lateral movement, supporting WinRM and SMB operations.
Security gaps in printer firmware and endpoint vetting are being increasingly targeted by attackers.
There is a rise in browser-based AI agent attacks, with security vendors emphasizing AI-first protection strategies.
AI-driven malware, such as LameHug, and the exploitation of AI tools for malicious purposes are emerging trends.
Cloud attack surface management and continuous threat exposure management (CTEM) are being discussed as alternatives to traditional vulnerability management.
Massistant, a mobile forensics tool used by Chinese law enforcement, is capable of collecting user information, files, and location data from confiscated devices.

Google has filed a lawsuit against operators of the BadBox 2.0 Android malware botnet, which infected 10 million devices and ran a global ad fraud scheme.
A settlement was reached in the class action lawsuit against Meta over the Cambridge Analytica incident; details remain undisclosed.
WeTransfer has retracted a clause that would have allowed the company to train AI models on user-uploaded files, following user backlash.
An Armenian national was extradited to the US and faces charges related to the Ryuk ransomware operation.