PerMIT: Securing Persistent Memory for LLM Agents via Machine-Checked Integrity Theorems and Scalable Community Trust

Key points

• Persistent memory is a largely undefended surface for injection and poisoning, and the same risk limits memory sharing among agents.1
• PerMIT mediates the release and use of untrusted environment-derived values through a finite memory interface, while preserving agent-driven multi-turn planning and tool choice.1
• Over this finite interface, we prove two integrity theorems that are also machine-checked in Lean 4, which together ensure that environment-derived memory reaches a sensitive action only if confirmed at the gate and influences the agent's behavior only through what the gate displays.1
• Under explicit assumptions, we prove and corroborate in simulation a co-scaling property: as the community grows, the risks of promoting poison and of wrongly evicting benign knowledge fall, while per-user verification effort does not rise.1

Sources (1)

• [1]PerMIT: Securing Persistent Memory for LLM Agents via Machine-Checked Integrity Theorems and Scalable Community Trust
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 20, 11:31 PM

  We present PerMIT, a memory architecture with provable integrity guarantees under an explicit trust assumption: only user-typed inputs and values confirmed at a gate are trusted.

  Persistent memory is a largely undefended surface for injection and poisoning, and the same risk limits memory sharing among agents.

Extractive summary: sentences quoted from the sources.