Ant Group unveils AI safety models for agents and multimodal systems

Ant Group’s AI Safety Lab open-sourced SingGuard-NSFA, a guardrail model for autonomous agents.

The model is designed to catch prompt injection, data theft, malicious code execution, resource abuse, and permission misuse before an agent acts.

Ant also shared more details on SingGuard, a multimodal safety model.

The release highlights growing demand for pre-action guardrails as AI agents become more autonomous.