Help Net SecurityNews
AI agents inherit employee access, and nobody's watching how they use it (opens helpnetsecurity.com)
Nudge Security's AI Adoption in Practice 2026 report, based on customer telemetry, finds employee-built AI agents (Rovo, Zapier, Claude Managed Agents, ChatGPT Agents) often run on their creator's OAuth grants or API keys. This leaves unmonitored access to company data, muddied audit trails and OAuth links that bypass inspection. No new attack is described.
- Category
- AI-Targeted