Metraipolis › archive › dc_66a9901afee7

Environment Steering: Using Data Flow Control to Improve Agent Utility and Safety — arXiv:2609.35807

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35807
License (as archived)
CC0-1.0
Archived
2026-10-07T16:24:40.583Z
SHA-256 of the archived text
37f1c0c2df49456cf07e7e15df39ac63eac5ddb66c86b0e5afa5fa0424acf646
Stable URL of this page
https://metraipolis.com/documento/dc_66a9901afee7

How to cite

Environment Steering: Using Data Flow Control to Improve Agent Utility and Safety — arXiv:2609.35807 — arXiv (en), https://arxiv.org/abs/2609.35807 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_66a9901afee7 (Metraipolis archive dc_66a9901afee7, sha256 37f1c0c2df49456cf07e7e15df39ac63eac5ddb66c86b0e5afa5fa0424acf646)
BibTeX
@misc{dc66a9901afee7,
  title = {Environment Steering: Using Data Flow Control to Improve Agent Utility and Safety — arXiv:2609.35807},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35807},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_66a9901afee7; sha256 37f1c0c2df49456cf07e7e15df39ac63eac5ddb66c86b0e5afa5fa0424acf646},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35807. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM agents can make unsafe tool calls even when instructed to behave safely. Existing defenses constrain agents before execution, modify tool inputs/outputs, or rely on LLM judges; these approaches may depend on model behavior or block unsafe actions without helping the agent recover. We argue that the execution environment should instead enforce safety as the agent runs and steer it toward safe alternatives when violations occur---we call this Environment Steering. We implement this by modeling the agent and harness execution state as database tables, track the record-level data flows, and check these data flows against declarative policies during runtime. When violations are detected, policy- and context-specific feedback steers the agent toward safe trajectories. On AgentDyn, this enables the agent to improve task success rate over no-defense while achieving 0% attack success rate.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Databases (cs.DB).
Pagina del lavoro: https://arxiv.org/abs/2609.35807 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T15:54:53.708Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).