Mercor says it was hit by cyberattack tied to compromise LiteLLM

a day ago

Mercor, an AI recruiting startup, confirmed it was affected by a security incident related to a supply chain attack via the open source project LiteLLM.
The hacking group TeamPCP compromised LiteLLM, with extortion group Lapsus$ claiming to have targeted Mercor and accessed its data, though the connection is unclear.
Mercor, founded in 2023, works with companies like OpenAI and Anthropic to train AI models using domain experts, with a valuation of $10 billion and over $2 million in daily payouts.
The company moved quickly to contain the incident, launching an investigation with third-party forensics experts and communicating with customers and contractors.
Lapsus$ shared a sample of allegedly stolen data, including Slack and ticketing references, plus videos of AI system conversations with contractors.
The LiteLLM compromise involved malicious code in its open source package, discovered last week and removed within hours, affecting many companies due to its high usage.
LiteLLM has updated compliance processes post-incident, switching from Delve to Vanta for certifications, but the full impact and data exposure are still under investigation.

Hasty Briefsbeta