🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Discover how multimodal #AI agents are vulnerable to data exfiltration through hidden prompt injections. Read the full report on our blog and proactively ensure the security of your AI tools:⬇️ research.trendmicro.com/4jPvipT

Earth Kurma’s use of #keyloggers like KMLOG for credential theft highlights the need for advanced monitoring and protection strategies. Read the full report on our blog:⬇️ research.trendmicro.com/3RZHPuZ

#PLeak exposes sensitive data like trade secrets and internal rules from LLMs. Organizations must heighten their vigilance against prompt leakage attacks. Learn how to protect your systems in our blog: ⬇️ research.trendmicro.com/4iFh49O

The Agenda #ransomware group has been observed using #SmokeLoader and a new loader, which we named #NETXLOADER. Due to its stealthy behavior, the new loader poses an increased risk of data theft and device compromise. Understand its threat: ⬇️ research.trendmicro.com/4jKcTed

Threat actors leverage #TikTok's vast reach with videos that instruct users to execute malicious commands under the guise of software activation. This highlights the importance of reinforcing security awareness. Learn more about this evolving threat:⬇️ research.trendmicro.com/4kmL4sd

FOG #ransomware is being distributed by cybercriminals trolling users by abusing the name of the Department of Government Efficiency (DOGE), or individuals connected to it. Discover more in our analysis: ⬇️ research.trendmicro.com/3EhznE8

Running Python code in an isolated sandbox isn’t enough to ensure secure execution. In this report, we highlight the need for robust monitoring, input validation, and resource controls. Read part two of our eight-part series on AI agents and #LLMs:⬇️ research.trendmicro.com/3EMhv4t

Read about the implications of weak cloud credentials on #PCManager's supply chain and learn the best practices in our blog. Proactive security begins by understanding this issue:⬇️ research.trendmicro.com/42wIMjb

Our research uncovers the methods North Korean cybercriminals, such as Void Dokkaebi, use to crack cryptocurrency wallets. Arm your SOC with the latest intelligence. Read more:⬇️ research.trendmicro.com/voiddokkaebi

AI agents face evolving threats, including prompt injection, where attackers manipulate AI outputs, and code execution, potentially leading to unauthorized system control. Discover security strategies in our ongoing series: ⬇️ research.trendmicro.com/4jLXt9v

Earth Kasha's latest campaign features an updated ANEL backdoor with a new command for executing Beacon Object Files (BOF) in memory. This continuous evolution underscores their advanced espionage tactics. Read more:⬇️ research.trendmicro.com/44eDJ9x

#BPFDoor attacks are targeting telecommunications, finance, and retail sectors in South Korea, Hong Kong, Myanmar, Malaysia, and Egypt. Be proactive in your defenses by understanding this threat. Read to learn more:⬇️ research.trendmicro.com/3GaWb94

The recent NVIDIA patch for the NVIDIA Container Toolkit vulnerability (CVE-2024-0132) doesn’t fully close security gaps. Discover the implications and what actions you may need to take in our research. Read more on our blog: ⬇️ research.trendmicro.com/43O6zxl

The #CREM solution calculates the #CRI to quantify an organization’s security risk on a scale from 1-100. A higher CRI means greater vulnerability. Learn how to reduce your CRI:⬇️ research.trendmicro.com/2025-Cyber-Ris…

Earth Ammit’s strategy centered around infiltrating the upstream segment of the drone supply chain. Our latest report details how they used compromised ERP software and custom malware to spy on drone manufacturers in Taiwan. Learn more: ⬇️ research.trendmicro.com/earthammit

What is VARGEIT, and how does the #EarthAlux #APT group weaponize it to infiltrate systems? This multi-channel #backdoor is a key component of their cyberespionage toolkit. Learn more about VARGEIT and other tools on our blog: ⬇️ research.trendmicro.com/4ca7Xw8

CrazyHunter’s application of the process killer tool, #ZammoCide, has five different iterations. This highlights a persistent evolution of the group’s operations. Learn more about #CrazyHunter in our blog: ⬇️ research.trendmicro.com/4jhqGJ2

Nation-state-aligned actors are increasingly using the Russian underground to achieve strategic objectives, combining financially driven attacks with political campaigns. Discover how these interactions are reshaping cyber operations in our report:⬇️ research.trendmicro.com/Russian_Underg…

Can an #LLM service become a cyberattack entry point? This report explores systemic risks from code execution in LLM-powered agents, highlighting attack vectors, security risks, and mitigations. Proactive security starts here:⬇️ research.trendmicro.com/3EMhv4t

Multimodal AI agents face unique data exfiltration threats from hidden instructions in images. Explore mitigation strategies and uncover these concealed risks on the blog:⬇️ research.trendmicro.com/4jPvipT