This paper evaluated Niyam-AI on 2,000 real-world agent scenarios from Agent-SafetyBench and compared it against three existing safety approaches: NeMo Guardrails, Meta's Llama Prompt Guard 2, and OpenAI's GPT-OSS-Safeguard.
NiyamAI is presented, an intent bound runtime guardrail architecture providing cryptographically verifiable execution integrity for autonomous agents and subjects its own enforcement mechanism to 18 adversarial vectors across six classes, disclosing two implementation vulnerabilities identified and remediated during de...
Aditya Katkar, Om Karkele, Kartik Mandhane et al.· 1 citation
Proof-of-Execution Memory (PoEM), which does not inspect memory at all, is proposed, which withstands attacks aimed at itself, adds microseconds of overhead, and works unchanged in a real LangChain agent.
This work builds a working instance on a hierarchical multi-agent system, runs it under benign and attacked conditions across five language models and two task domains, and measures how much of that warning rests on removable surface cues of the attack rather than on its distributed structure.
D. Arias, Dev Prashant Mistry, Ren Wang et al.· arXiv.org· 0 citations
What an agent can be trusted to establish is bounded by the strength of its feedback, and the central lesson is what an agent can be trusted to establish is bounded by the strength of its feedback.
The threat model in the HPC setting is defined, attack surfaces created by schedulers, shared storage, multi-project accounts, and scientific workflows are identified, and where current controls fall short is examined.
Preliminary evidence is found that a single sentence added to the developer's own prompt substantially improves disclosure, and a benchmark scored on attack success alone cannot rank agents on the risk a developer actually carries.
Aaron Su, Rong-Xing Lu· Inquiry@Queen's Undergraduat...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
