Xiang-Fan Wu

The account motivates complementary defenses: strengthen resistance and recovery alongside prevention of spontaneous deviations, and audit and restrict unintended communication paths that can turn local failures into collective loss of control.

These results show how a latent loss of control can become an external violation: the task objective remains intact, but an executable opportunity can turn a missing control boundary into consequential action.

Indirect prompt injection can compromise AI systems like DeepSeek Harness, with attack success rates reaching up to 25.5% under certain conditions.

Extreme fidelity loss reveals critical vulnerabilities in long-horizon tasks that standard accuracy metrics overlook.

SkillJack reveals that self-evolving agents can unknowingly incorporate malicious skills, making traditional safety measures ineffective against persistent threats.

SkillSentry reveals that dynamic testing can uncover harmful agent behaviors that static analysis fails to detect, achieving unprecedented accuracy in safety evaluations.

SafeFlow reveals that multi-agent systems can obscure malicious intent through task decomposition, but a semantic information-flow approach can effectively counteract this vulnerability.

ShadowProbe uncovers hidden algorithmic risks in codebases, revealing vulnerabilities that traditional methods miss, with significant implications for software reliability.

AI-Infra-Guard reveals that a unified security framework can effectively address the diverse attack surfaces of AI agents, making it a game-changer for AI safety.