Show HN: AgentTrust – Zero-trust security and human approvals for AI agents
AI summary
Editorial policyAn enterprise zero-trust security gateway for AI agents featuring real-time prompt injection defense, action scoping, human-in-the-loop approvals, post-quantum Merkle auditing, and EU AI Act compliance.
Key points
- •Implements real-time prompt injection defense and action scoping for AI agents.
- •Includes human-in-the-loop (HITL) approval processes and post-quantum Merkle audit capabilities.
- •Designed as an enterprise governance gateway to support compliance with the EU AI Act.
#Security & Misuse#AI Agents#Regulation, Ethics & Society#AgentTrust