Microsoft packages hit with credential-stealing malware for second time

Seventy-three Microsoft open source packages were compromised late last week with malware that steals credentials from cloud services and developer tools. The malicious code activates when opened in AI coding agents.

Automated systems on GitHub blocked the packages after detecting the threat. GitHub disabled them citing a terms of service violation rather than labeling them malicious. Microsoft sent an email on Monday stating it had temporarily removed some repositories while investigating potential malicious content. This marks the second such incident involving a Microsoft account in recent months. The malware, known as Miasma, uses a 28-kilobyte payload linked to threat actor TeamPCP. It targets credentials for AWS, Azure, GCP, Kubernetes, and password managers before spreading laterally. The same GitHub account was used in a May compromise of the DurableTask Python SDK. Security researchers noted the attack bypasses traditional detection by generating unique encrypted payloads for each infection.

관련 기사

Illustration of a hacker exploiting Meta's AI chatbot to hijack Instagram accounts by changing email addresses and bypassing security.
AI에 의해 생성된 이미지

Meta patches ai chatbot flaw used to hijack instagram accounts

AI에 의해 보고됨 AI에 의해 생성된 이미지

Hackers exploited Meta's AI support chatbot to take over Instagram accounts by tricking it into changing associated email addresses. The vulnerability allowed password resets without two-factor authentication after matching locations via VPN. Meta resolved the issue with an emergency patch on May 29.

GitHub was targeted in a significant cyber attack involving malware-laden commits. The Megalodon operation affected more than 5,000 repositories.

AI에 의해 보고됨

Microsoft has alerted users that hackers are targeting password reset processes to breach accounts. The activity is attributed to the group Storm-2949.

A new report indicates that most companies have released software containing known security flaws. The problem is especially pronounced with AI-created code, which exceeds the speed of manual fixes.

이 웹사이트는 쿠키를 사용합니다

사이트를 개선하기 위해 분석을 위한 쿠키를 사용합니다. 자세한 내용은 개인정보 보호 정책을 읽으세요.
거부