AITechNodeabout 2 hours ago
Ant Group open-sources AI safety models
1 min read
Ant Group's AI Safety Lab open-sourced SingGuard-NSFA, a safety guardrail model for autonomous agents, and disclosed SingGuard, a multimodal safety model. SingGuard-NSFA detects risks like prompt injection and malicious code execution across 185 scenarios and 133 languages, with nearly 100,000 evaluation samples. The model is available in 0.8B, 2B, 4B, and 9B versions, making a risk judgment in about 50 milliseconds.
Level
Hype check
Tap to vote and see what everyone thinks.
