Agent Memory Is a Surface for Endogenous Authorization Laundering

arXiv:2609.01836 · cs.CR, cs.AI · Submitted 2026-09-01 · Read on arXiv

cs.CR, cs.AI

Submitted: 2026-09-01

Updated: 2026-09-01

Code: https://github.com/tommasocerruti/eal-bench

Terminology

• Benchmark — 15× in this paper · explained in 2 episodes
• Provenance — 15× in this paper · explained in 4 episodes
• Ablation — 5× in this paper · explained in 1 episode
• Alignment — 4× in this paper · explained in 1 episode
• Calibration — 4× in this paper · explained in 7 episodes
• Fidelity — 4× in this paper · explained in 2 episodes
• LLM judge — 4× in this paper · explained in 2 episodes
• AI Agents — 3× in this paper · explained in 2 episodes
• Benchmarking — 3× in this paper · explained in 3 episodes
• Agent Memory — 2× in this paper · explained in 1 episode
• Robustness — 2× in this paper · explained in 5 episodes
• Abstraction — 1× in this paper · explained in 1 episode
• Actor — 1× in this paper · explained in 3 episodes
• Adversarial Attacks — 1× in this paper · explained in 1 episode
• Adversarial scenarios — 1× in this paper · explained in 1 episode
• Agent Architectures — 1× in this paper · explained in 1 episode
• Agent Harness — 1× in this paper · explained in 2 episodes
• Alignment Problem — 1× in this paper · explained in 1 episode
• Core — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Summer of Open AI Research — 1× in this paper
• Trace — 1× in this paper · explained in 2 episodes
• Transformer — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• CyberSecEval 2: A Wide-Ranging Cybersecurity Evaluation Suite for Large Language Models
• Defensive Refusal Bias: How Safety Alignment Fails Cyber Defenders
• STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• GLM-5: from Vibe Coding to Agentic Engineering
• Evaluation Awareness in Language Models Has Limited Effect on Behaviour
• Auditing Agent Harness Safety
• How Inference Compute Shapes Frontier LLM Evaluation
• How Agents Ask for Permission: User Permissions for AI Agents, from Interfaces to Enforcement
• Large Language Models Often Know When They Are Being Evaluated
• Why Do AI Agents Break Rules? How Framing, Context, and Social Signals Shape Compliance
• MemGPT: Towards LLMs as Operating Systems
• GateMem: Benchmarking Memory Governance in Multi-Principal Shared-Memory Agents
• Toolformer: Language Models Can Teach Themselves to Use Tools
• Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
• AI Sandbagging: Language Models can Strategically Underperform on Evaluations
• Memory Provenance Laundering in LLM Agents: A Non-Amplification Firewall for Persistent Memory
• MasDrift: Benchmarking Authorization Preservation Across Multi-Agent Architectures
• ReAct: Synergizing Reasoning and Acting in Language Models
• When Memory Becomes Authority: Benchmarking Authority Collapse at the Memory Consolidation Boundary

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs