‘GitLost’ vulnerability let GitHub’s AI workflows leak private repositories
A critical prompt injection flaw in GitHub's Agentic Workflows risked exposing sensitive private repository data.
Researchers at artificial intelligence security company Noma Security Inc. today disclosed a critical prompt injection vulnerability in GitHub Inc.'s new Agentic Workflows feature that allowed an unauthenticated attacker to siphon data from private code repositories by posting a ...
SiliconANGLE
Source

Generated insight by OnAirToday AI
The GitLost vulnerability highlights the precarious nature of agentic workflows when they operate with excessive system permissions. It serves as a stark reminder that integrating AI into software development pipelines requires strict isolation boundaries to prevent unauthorized data exposure.
Get the day's AI news in one short email. Free.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.