#Artificial Intelligence Security

Articles (3)

• 🇺🇸 AgentLAB: Benchmarking LLM Agents against Long-Horizon Attacks — 20/02/2026 [USA]

  arXiv:2602.16901v1 Announce Type: new Abstract: LLM agents are increasingly deployed in long-horizon, complex environments to solve challenging problems, but this expansion exposes them to long-horiz...

  Related: #Large Language Model Agents, #Long‑Horizon Attack Vectors, #Benchmark Development, #Multi‑Turn Interaction Vulnerabilities
• 🇺🇸 Protecting Language Models Against Unauthorized Distillation through Trace Rewriting — 18/02/2026 [USA]

  arXiv:2602.15143v1 Announce Type: new Abstract: Knowledge distillation is a widely adopted technique for transferring capabilities from LLMs to smaller, more efficient student models. However, unauth...

  Related: #Intellectual Property Rights in AI, #Model Knowledge Distillation, #Trace-Based Anti‑Distillation Methods, #Detection of Unauthorized Model Use
• 🇺🇸 Zombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing Injections — 18/02/2026 [USA]

  arXiv:2602.15654v1 Announce Type: cross Abstract: Self-evolving LLM agents update their internal state across sessions, often by writing and reusing long-term memory. This design improves performance...

  Related: #Long‑Term Memory in LLMs, #Persistent Malware/Attack Vectors, #Ethical Design of Autonomous Agents