Malicious LiteLLM Releases Tied to Trivy Hack May Have Exposed 2,100+ Organizations
- ID
- 13390
- Status
- summarized
- Published
- 12 Aug 2026, 4:04 PM
- Fetched
- 12 Aug 2026, 5:22 PM
- Provider
- The Hacker News
- Category
- security
- Original URL
- https://thehackernews.com/2026/08/malicious-litellm-releases-tied-to.html
- Source URL
- https://feeds.feedburner.com/TheHackersNews
Summary
- Score
- 7.5
- Created
- 12 Aug 2026, 5:22 PM
- Tags
- Audience
- developersvibe_codersai_agent_usersai_ml_learners
What happened
Two malicious LiteLLM releases (versions 1.82.7 and 1.82.8) were live on PyPI for ~40 minutes on March 24, 2026, containing credential-stealing code that harvested cloud keys, SSH keys, Kubernetes tokens, and database passwords. CloudSEK obtained ~434,000 captured files mapping potential exposure to 2,500+ organizations (including NVIDIA, Cisco, Deloitte, Volkswagen), and published a public lookup tool. The FBI warned in a July advisory that stolen credentials may be weaponized long after the initial compromise.
Why it matters
If you installed LiteLLM from PyPI on March 24, 2026 (especially between 10:39–16:00 UTC), treat your CI/CD secrets as compromised and rotate cloud keys, SSH keys, Kubernetes tokens, and database passwords immediately—do not wait for proof of misuse. Check CloudSEK's public lookup tool by org name or domain to assess exposure.
Discussion angle
How many teams using LiteLLM as their AI gateway even knew this happened in March, and what does it say about trusting PyPI packages in CI pipelines for AI infrastructure?