List your company

Prompt Injection Attack on InjecAgent

0ASR @ 1 AttemptDirect
-4235077Mar 13, 2026

Evaluation results

MethodLinks
DirectTarget Model=Meta-SecAlign-8B, Attack Category=Static Attacks2026.030—
DirectTarget Model=GPT-5-nano, Attack Category=Static Attacks2026.030—
EnhancedTarget Model=GPT-5-nano, Attack Category=Static Attacks2026.030—
EnhancedTarget Model=Meta-SecAlign-8B, Attack Category=Static Attacks2026.031—
PAIRTarget Model=Meta-SecAlign-8B, Attack Category=Search-Based Attacks2026.031—
DirectTarget Model=GPT-4.1-nano, Attack Category=Static Attacks2026.031—
PAIRTarget Model=GPT-5-nano, Attack Category=Search-Based Attacks2026.031—
DirectTarget Model=GPT-4o-mini, Attack Category=Static Attacks2026.032—
EnhancedTarget Model=GPT-4.1-nano, Attack Category=Static Attacks2026.032—
TAPTarget Model=Meta-SecAlign-8B, Attack Category=Search-Based Attacks2026.033—
EnhancedTarget Model=GPT-4o-mini, Attack Category=Static Attacks2026.033—
StrategyTarget Model=Meta-SecAlign-8B, Attack Category=Search-Based Attacks2026.034—
Vanilla GRPOTarget Model=Meta-SecAlign-8B, Attack Category=RL-Based Attacks2026.03611
TAPTarget Model=GPT-5-nano, Attack Category=Search-Based Attacks2026.038—
Vanilla GRPOTarget Model=GPT-5-nano, Attack Category=RL-Based Attacks2026.031124
StrategyTarget Model=GPT-5-nano, Attack Category=Search-Based Attacks2026.0318—
PAIRTarget Model=GPT-4o-mini, Attack Category=Search-Based Attacks2026.0324—
PAIRTarget Model=GPT-4.1-nano, Attack Category=Search-Based Attacks2026.0332—
StrategyTarget Model=GPT-4o-mini, Attack Category=Search-Based Attacks2026.0338—
TAPTarget Model=GPT-4o-mini, Attack Category=Search-Based Attacks2026.0340—
Vanilla GRPOTarget Model=GPT-4o-mini, Attack Category=RL-Based Attacks2026.034660
TAPTarget Model=GPT-4.1-nano, Attack Category=Search-Based Attacks2026.0354—
Vanilla GRPOTarget Model=GPT-4.1-nano, Attack Category=RL-Based Attacks2026.036175
StrategyTarget Model=GPT-4.1-nano, Attack Category=Search-Based Attacks2026.0365—
RL-HammerTarget Model=GPT-5-nano, Attack Category=RL-Based Attacks2026.038696
RL-HammerTarget Model=GPT-4o-mini, Attack Category=RL-Based Attacks2026.0394100
RL-HammerTarget Model=GPT-4.1-nano, Attack Category=RL-Based Attacks2026.0395100
PISmithTarget Model=GPT-5-nano, Attack Category=RL-Based Attacks2026.0395100
RL-HammerTarget Model=Meta-SecAlign-8B, Attack Category=RL-Based Attacks2026.0398100
PISmithTarget Model=Meta-SecAlign-8B, Attack Category=RL-Based Attacks2026.0399100
PISmithTarget Model=GPT-4o-mini, Attack Category=RL-Based Attacks2026.0399100
PISmithTarget Model=GPT-4.1-nano, Attack Category=RL-Based Attacks2026.03100100