Hugging Face Daily Papers · · 3 min read

Agent Memory Is a Surface for Endogenous Authorization Laundering

Mirrored from Hugging Face Daily Papers for archival readability. Support the source by reading on the original site.

.</p>\n","updatedAt":"2026-09-03T00:26:05.602Z","author":{"_id":"65044901221035b9ee187178","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65044901221035b9ee187178/o2LTLLIoT3T17Y4ZZZy_x.jpeg","fullname":"Tommaso Cerruti","name":"Cerru02","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":5,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"fr","probability":0.32275810837745667},"editors":["Cerru02"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/65044901221035b9ee187178/o2LTLLIoT3T17Y4ZZZy_x.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2609.01836","authors":[{"_id":"6a98bc61fea818274321fdda","user":{"_id":"65044901221035b9ee187178","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65044901221035b9ee187178/o2LTLLIoT3T17Y4ZZZy_x.jpeg","isPro":false,"fullname":"Tommaso Cerruti","user":"Cerru02","type":"user","name":"Cerru02"},"name":"Tommaso Cerruti","status":"claimed_verified","statusLastChangedAt":"2026-09-03T00:45:04.498Z","hidden":false},{"_id":"6a98bc61fea818274321fddb","name":"Mika Okamoto","hidden":false},{"_id":"6a98bc61fea818274321fddc","name":"Ansel Kaplan Erol","hidden":false}],"publishedAt":"2026-09-01T00:00:00.000Z","submittedOnDailyAt":"2026-09-02T00:00:00.000Z","title":"Agent Memory Is a Surface for Endogenous Authorization Laundering","submittedOnDailyBy":{"_id":"65044901221035b9ee187178","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65044901221035b9ee187178/o2LTLLIoT3T17Y4ZZZy_x.jpeg","isPro":false,"fullname":"Tommaso Cerruti","user":"Cerru02","type":"user","name":"Cerru02"},"summary":"Long-running LLM agents rely on persistent memory to carry state across interactions, including permissions, restrictions, and revocations. When memory misrepresents this evolving authorization state, the agent's own records can grant authority that the underlying history never permitted, resulting in misaligned behavior without any external attacks.\n We term this failure endogenous authorization laundering, where spurious permissions written into memory lead to unauthorized actions as their provenance is washed away. We then introduce EAL-Bench, which measures how accurately persistent memory preserves evolving authorization state and whether errors propagate to downstream unauthorized actions.\n We evaluate five LLMs as memory writers and two as executors across procurement, cybersecurity, and finance. We find that under incremental memory updates, writers create false authority for up to 50.2% of unauthorized requests; once false authority is present, executors act on it in 98.6% of trials. Two safeguards, requiring stored permissions to be backed by valid source events, and tracking permission changes through bounded event sourcing, substantially reduce laundering, but both also reject more legitimate actions, exposing a safety-utility tradeoff. Persistent memory is therefore not merely a performance component, but a part of an LLM agent's effective authorization policy.","upvotes":3,"discussionId":"6a98bc62fea818274321fddd","ai_summary":"Persistent memory errors in long-running LLM agents can falsely grant authority, leading to unauthorized actions, and safeguards reduce this at a cost to legitimate use.","ai_keywords":["LLM agents","persistent memory","endogenous authorization laundering","EAL-Bench","bounded event sourcing","authorization state","memory writers"],"ai_summary_model":"thinkingmachines/Inkling-Small","organization":{"_id":"60347fbbe8149a962412a689","name":"EleutherAI","fullname":"EleutherAI","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/1614054059123-603481bb60e3dd96631c9095.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"65044901221035b9ee187178","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65044901221035b9ee187178/o2LTLLIoT3T17Y4ZZZy_x.jpeg","isPro":false,"fullname":"Tommaso Cerruti","user":"Cerru02","type":"user"},{"_id":"63ac5701c21e60a3e9b58aa7","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63ac5701c21e60a3e9b58aa7/g6EX7diOpuA94R2ab-rZC.png","isPro":true,"fullname":"Dipankar Sarkar","user":"dipankarsarkar","type":"user"},{"_id":"69bd161dc72e20286157e223","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/ly1ld8fcRubxmyeef4lzN.png","isPro":false,"fullname":"小林陽太","user":"victking72","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"60347fbbe8149a962412a689","name":"EleutherAI","fullname":"EleutherAI","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/1614054059123-603481bb60e3dd96631c9095.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2609/2609.01836.md","query":{}}">
Papers
arxiv:2609.01836

Agent Memory Is a Surface for Endogenous Authorization Laundering

Published on Sep 1
· Submitted by
Tommaso Cerruti
on Sep 2
Authors:

Abstract

Persistent memory errors in long-running LLM agents can falsely grant authority, leading to unauthorized actions, and safeguards reduce this at a cost to legitimate use.

Long-running LLM agents rely on persistent memory to carry state across interactions, including permissions, restrictions, and revocations. When memory misrepresents this evolving authorization state, the agent's own records can grant authority that the underlying history never permitted, resulting in misaligned behavior without any external attacks. We term this failure endogenous authorization laundering, where spurious permissions written into memory lead to unauthorized actions as their provenance is washed away. We then introduce EAL-Bench, which measures how accurately persistent memory preserves evolving authorization state and whether errors propagate to downstream unauthorized actions. We evaluate five LLMs as memory writers and two as executors across procurement, cybersecurity, and finance. We find that under incremental memory updates, writers create false authority for up to 50.2% of unauthorized requests; once false authority is present, executors act on it in 98.6% of trials. Two safeguards, requiring stored permissions to be backed by valid source events, and tracking permission changes through bounded event sourcing, substantially reduce laundering, but both also reject more legitimate actions, exposing a safety-utility tradeoff. Persistent memory is therefore not merely a performance component, but a part of an LLM agent's effective authorization policy.

Community

Paper author Paper submitter about 2 hours ago
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images

· Sign up or log in to comment

Get this paper in your agent:

hf papers read 2609.01836
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper

No model linking this paper

Cite arxiv.org/abs/2609.01836 in a model README.md to link it from this page.

Datasets citing this paper

No dataset linking this paper

Cite arxiv.org/abs/2609.01836 in a dataset README.md to link it from this page.

Spaces citing this paper

No Space linking this paper

Cite arxiv.org/abs/2609.01836 in a Space README.md to link it from this page.

Collections including this paper

No Collection including this paper

Add this paper to a collection to link it from this page.

Discussion (0)

Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.

Sign in →

No comments yet. Sign in and be the first to say something.