Crimson AI NewsA CrimsonLingua Network service
EN ع
← Back to news
Research paper Hugging Face

Bounded Agents: New Authorization Model Cuts Multi-Agent Data Theft to Zero

AI By Crimson AI Hugging Face Papers 20 August 2026 · 00:00 6 views
Share: X Telegram

Researchers propose the Agentic Principal Chain (APC), an authorization architecture that tracks delegated authority across multi-agent AI systems, reducing data exfiltration from 75-100% to 0% in benchmarks.

Bounded Agents: New Authorization Model Cuts Multi-Agent Data Theft to Zero

Key points

A new research paper from Hugging Face introduces the Agentic Principal Chain (APC), an authorization model designed to secure multi-agent AI systems against delegation abuses and harmful action combinations. The work addresses a critical gap: LLM agents typically have static permissions set at session start, and each request is evaluated independently, ignoring prior actions.

The authors argue that prompt injection attacks are fundamentally an authorization problem, not just a model robustness issue. An agent may combine individually permitted actions into a prohibited outcome or delegate authority to a sub-agent without proper limits. APC tackles this by tracking delegated authority from one principal to the next, evaluating each request against accumulated session state using six authorization checks.

Key features include carrying forward and restricting delegated scope and budgets, and using composition closure to check requests against prior actions, preventing prohibited combinations. Enforcement is done outside the model, ensuring decisions are not influenced by the model itself.

The paper proves two properties for APC implementations: Blast Radius Monotonicity and Composition Soundness (the latter limited to prohibited combinations under a complete restriction set and serialized admission). In evaluations across 3,154 instances, APC reduced AgentDojo exfiltration from 75-100% to 0% across all four domains, blocked all 544 InjecAgent data-stealing cases, and cut destruction from 38.6% to 4.0% and manipulation from 90.5% to 12.1%.

Authorization latency was just 0.24 ms at the 99th percentile on an idle host. Utility dropped by 8.6 and 13.9 percentage points in two AgentDojo settings. The implementation, evaluation tools, and data are publicly available.

MetricBefore APCAfter APC
AgentDojo exfiltration (all domains)75-100%0%
InjecAgent data-stealing cases blocked544/544
Destruction rate38.6%4.0%
Manipulation rate90.5%12.1%
Authorization latency (99th percentile)0.24 ms
Source
Hugging Face · Hugging Face Papers
Related news
Research paper
Hugging Face 29 Aug 2026

Hugging Face Audit: 110 of 124 AI Evaluations Fail to Support Their Claims

A new commit-bound census of 124 Inspect Evals units reveals that 110 stop before deterministic inference due to missing historica...

4
Research paper
Hugging Face 29 Aug 2026

Aphanta: New Framework Diagnoses When Image Editing Boosts Multimodal Reasoning

Hugging Face researchers introduce Aphanta, a diagnostic framework that evaluates when image-editing intermediates improve multimo...

5
Research paper
Hugging Face 29 Aug 2026

EditaLive! Enables Real-Time Character Video Editing for Live Streaming

Hugging Face researchers introduce EditaLive, a framework for real-time human-centric video editing in live streams, achieving sta...

4