Zhiping Zhang, Franklin Mingzhe Li, Zihao Huang, Tianshi Li · 2026 · arXiv
DOI: 10.48550/arXiv.2610.27548 · open-access PDF
Ketika agen LLM bertindak atas nama pengguna, diharapkan untuk bertindak dalam lingkup otorisasi yang diberikan pengguna. Evaluasi agen yang ada sebagian besar mengukur penyelesaian tugas atau konsekuensi berbahaya dan karena itu dapat mengabaikan tindakan yang melanggar delegasi ini tanpa menghasilkan keamanan konvensional atau kerusakan privasi. Kami memperkenalkan AuthorizationBench, benchmark dari penyesuaian otorisasi yang didasarkan pada kasus dunia nyata yang didokumentasikan.
Machine translation (Meta M2M100); the English original follows.
When an LLM agent acts on behalf of a user, it is expected to act within the authorization scope the user grants. Existing agent evaluations largely measure task completion or harmful consequences and can therefore overlook actions that violate this delegation without producing conventional safety or privacy harms. We introduce AuthorizationBench, a benchmark of authorization alignment grounded in documented real-world cases. A scalable construction pipeline produces 286 executable tasks spanning situational triggers, the presence or absence of direct harmful consequences, and ten application
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.27548, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange