AI Coding Tools Now a Prime Target for Threat Actors, Google Warns

Google Threat Intelligence Group reports a surge in attacks targeting AI-assisted development and proprietary AI data, highlighting UNC6780’s large-scale open-source supply-chain compromises using Dustmaker credential stealer (exfiltrating GitHub Actions tokens and implanting malicious files into AI assistant workspaces) and UNC6508’s espionage targeting AI research; the report also documents mass credential harvesting, data-extortion incidents, and increasing adversary experimentation with agentic AI tools.