Sponsored
SECURITY
New genAI apps are popping up every other day. You need a way to discover, secure, and govern genAI usage [4] — ideally, one that's more scalable than chasing people over email.
Parse notes
Sponsored
New genAI apps are popping up every other day. You need a way to discover, secure, and govern genAI usage [4] — ideally, one that's more scalable than chasing people over email.
American Water, the largest U.S. water utility, reported a cyberattack on Thursday. The company paused billing and shut down certain systems but believes facilities and operations weren't affected. It is working with law enforcement to investigate the incident's scope.
A new Mirai variant called GorillaBot launched 300,000 DDoS attacks last month, affecting 20,000 organizations worldwide. It employs 19 attack methods, primarily UDP and TCP ACK Bypass floods. The botnet, which supports multiple architectures, targeted organizations in 113 countries, with China and the US most affected.
5.11 Tactical, which mashes tactical equipment that is favored by local and national law enforcement agencies, suffered a data breach. In a filing with the Maine Attorney General, 5.11 Tactical disclosed that 27,742 individuals who purchased products online between July 12 to August 22 may have had their information exposed. Exposed information includes full names, email addresses, payment card numbers, payment card expiration dates, and payment card security codes.
The Phishing-as-a-Service platform Sniper Dz has been detected using over 140,000 unique phishing websites. It uses various techniques to hide its malicious activity such as utilizing proxymesh[.]com to hide backend servers, obfuscating template code, abusing legitimate SaaS platforms, and employing malicious redirects. Sniper Dz offers these services free of charge on its platform and also provides a centralized infrastructure for exfiltrating credentials and tracking victims.
Websites can employ CORS to allow for sharing of data to a different website, bypassing same-origin restrictions. This blog post presents a collection of case studies that showcase CORS misconfigurations that can be exploited such as arbitrary reflected origin, only checking the beginning of the Origin header, trusting related subdomains or localhost, and other more complicated techniques. Advice to pentesters hunting for CORS vulnerabilities includes making sure to scan all endpoints, considering related domains, applying context to findings, and testing for subdomain reflection.
This is a series of articles that describe how Docker Container Security works, how to exploit vulnerabilities as well as how to fix and secure your Docker deployment. Another series on Linux Privilege Escalation as well as some fundamental understanding of Docker are prerequisites for this course.
Lynis is a security auditing tool for systems based on UNIX like Linux, macOS, BSD, and others. It performs an in-depth security scan and runs on the system itself. The primary goal is to test security defenses and provide tips for further system hardening.
Tamnoon combines purpose-built AI technology for cloud remediation, guided by cloud security experts, to safely reduce critical threat exposures.
Ghostport is a tool that misleads port scanners by responding with convincing signatures.
Ukrainian-linked hackers reportedly conducted an "unprecedented" cyberattack on Russian state broadcaster VGTRK, disabling online broadcasting and internal services, on Putin's birthday. The Kremlin expressed support and is investigating the incident.
Apple has released macOS 15.0.1, addressing compatibility issues with third-party security software that arose in the initial Sequoia release. The update reportedly fixes networking problems that affected products from CrowdStrike and Microsoft.
Most vulnerability management tools only handle the simplest portion of vulnerability management - finding vulnerabilities. Most tools don't have enough context to provide helpful remediation advice for practitioners to take action on their findings. The future of vulnerability management tools will include a set of AI agents that constantly build up context and state of the environment to not only find vulnerabilities, but pinpoint everywhere that they're in use and how they could be fixed.
The Chrome Security team describes how they evaluate and prioritize security improvements based on how they reduce potential harm to users and increase costs for attackers, considering factors like bug reliability, ease of exploitation, and impact across Chrome's large user base.
Microsoft, Adobe, and Apple have released security updates to fix numerous vulnerabilities, including active attacks on Windows and other software.
Lua malware has been discovered in fake game cheats and is spread globally through techniques like SEO poisoning.