The Trojan Hippo Bench is introduced, a dynamic evaluation framework for persistent memory attacks and memory-layer defenses, comprising an OpenEvolve-based adaptive red-teaming benchmark that stress-tests defenses and memory backends against continuously refined attacks, and a capability-aware security-utility analysi...
Debeshee Das, Julien Piet, D. Kaviani et al.· 12 citations· ⚡1
This work identifies a concrete failure of trace integrity in agent infrastructure which can be used to conceal misaligned behaviors like scheming or sabotage, and advises practitioners to ensure trace logging happens through an independent interception mechanism outside of the agent's control, preserving trace integri...
Jeremy Qin, David Schmotz, Derck W. E. Prinzhorn et al.· 0 citations
It is found that GPT-6 Astra's low evasion rate comes with overrefusal, as it frequently abandons otherwise solvable tasks under a denial-of-service prompt injection.
David Schmotz, Derck W. E. Prinzhorn, Luca Beurer-Kellner et al.· 1 citation
Several open research questions are identified, including how to efficiently record incidents and how to determine whether vulnerabilities and incidents generalize, and privacy requirements are summarized and research directions for the secure and trustworthy deployment of AI agents are outlined.
Anastasia Pustozerova, E. Bagdasarian, Luca Beurer-Kellner et al.· 0 citations
An architectural vulnerability is identified that circumvents anti-distillation mechanisms, allowing adversaries to extract a proprietary model's reasoning, as well as proposing concrete cryptographic and system-level mitigations to secure client-side reasoning.
Alexander Panfilov, David Schmotz, I. Shumailov et al.· 9 citations· ⚡3
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
