Amazon Bedrock environments, which support large language models (LLMs) for artificial intelligence tools, have been accessed by threat actors using exposed AWS access keys. The breaches, some of which exploited the LLMs for inappropriate interactions, were made possible by disregarding security best practices, including the misuse of long-term keys. Amazon has assured that no customers were endangered and has taken additional action to restrict compromised accounts.
From a Teams Call to a Ransomware Threat: Matanbuchus 3.0 MaaS Levels Up – Morphisec
From a Teams Call to a Ransomware Threat: Matanbuchus 3.0 MaaS Levels Up Morphisec
