An Australian advisory details how compromised credentials, sessions, vulnerable applications and third-party access can expose AI services, and recommends practical controls for accounts, keys, monitoring and incident response.
A joint agency advisory says WaterPlum, also known as Contagious Interview, uses fake recruiting and coding tests to infect IT professionals and steal credentials and cryptocurrency. It details the activity, a Japanese hiring case and defenses.
The guidance explains how agentic AI harnesses connect LLMs to organizational data and tools, creating security and operational risks. It recommends least privilege, human approval, monitoring, logging and phased deployment to manage those risks.
A joint advisory says Russian state-supported LAUNDRY BEAR has exploited CVE-2025-66376 in Zimbra webmail since 2025, using the Ulej payload and Flowerbed infrastructure to collect emails and account data.
The advisory explains how crypter services obfuscate and encrypt malware to evade detection, including their builder/stub model and counter-antivirus checks, and recommends EDR/HIPS, application control, hardened antivirus settings, and runtime behavioral analysis.
