Thousands of malicious AI skills found capable of stealing data

Source

Where: Global
When: Published July 8, 2026
What happened: Help Net Security covered ESET research finding more than 25,000 suspicious AI skills and over 3,000 malicious ones across a large sample of agent skills. These skills can support command execution, file access, credential loading, code injection and obfuscation, creating a new supply-chain risk for AI-agent ecosystems.