Agent Safety Evaluation on Agent-SafetyBench
72.3Agent-SafetyBench Scoregpt-4o + GBT-SE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| gpt-4o + GBT-SEBase Model=gpt-4o, Safety/Security Framework=GBT-SE2026.01 | 72.3 | — | — | |
| llama-3-8b + GBT-SEBase Model=llama-3-8b, Safety/Security Framework=GBT-SE2026.01 | 70.8 | — | — | |
| gpt-4o + GBT-BasicBase Model=gpt-4o, Safety/Security Framework=GBT-Basic2026.01 | 60.2 | — | — | |
| gpt-4o + Global guardrail onlyBase Model=gpt-4o, Safety/Security Framework=Global guardrail only2026.01 | 56.8 | — | — | |
| llama-3-8b + GBT-BasicBase Model=llama-3-8b, Safety/Security Framework=GBT-Basic2026.01 | 55.4 | — | — | |
| llama-3-8b + Global guardrail onlyBase Model=llama-3-8b, Safety/Security Framework=Global guardrail only2026.01 | 50.4 | — | — | |
| gpt-4o (native)Base Model=gpt-4o, Safety/Security Framework=native2026.01 | 44.2 | — | — | |
| llama-3-8b (native)Base Model=llama-3-8b, Safety/Security Framework=native2026.01 | 20.1 | — | — | |
| CaMeL2026.05 | — | 78.3 | 28.5 | |
| FIDES2026.05 | — | 72 | 42.1 | |
| NoDefense2026.05 | — | 0 | 100 | |
| PACT2026.05 | — | 74.7 | 39.3 |