Unrevoked LiteLLM key used to publish credential-stealing PyPI package
Original title: Popular AI tool LiteLLM says unrevoked key allowed hackers to target users
Microsoft said the malware ran on every Python startup, and stolen AWS keys were checked against Amazon AI models.
LiteLLM said an internal master key exposed in May and never revoked let an unidentified attacker take control of its GitHub repositories. Last week the attacker published PyPI package 1.107.0 with credential-harvesting malware that ran whenever Python started. Wiz said the package was downloaded about 49,000 times and stolen cloud logins were tested within hours, including against Amazon AI models. LiteLLM told anyone who installed 1.107.0 on an internet-connected computer to assume credentials were stolen.
Unconfirmed: No named organization has confirmed a breach, and the attacker is unidentified; download figures come from Wiz.
- The exposed key was in a configuration file for an internal AI-agent tool.
- The malicious package stayed on PyPI for about two hours.
- Within about four and a half hours, stolen AWS logins were tested.
- This is LiteLLM’s second PyPI supply-chain compromise this year.
- March malicious releases were later linked to TeamPCP; two men were charged.
Impact: Cloud and other credentials stolen from machines that installed the package.
| Product | Affected |
|---|---|
| LiteLLM | 1.107.0 |
- 1.Assume credentials were stolen if version 1.107.0 was installed on an internet-connected computer.
- 2.Change every password and access key that the affected computer could reach.
From the original article
2 min read at therecord.mediaLiteLLM, an open-source tool widely used by artificial intelligence developers, said a key accidentally exposed in May was behind a supply-chain attack that is believed to have hit tens of thousands of users last week.
therecord.media
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.