Agent Safety Is Undecidable When Coupled with General Computing Devices
arXiv — cs.AI ·
What to know
A new arXiv paper (2611.25325) shows that coupling a fixed language model with a general computing device makes agent safety undecidable: no algorithm can predict whether an agent will run a semantically equivalent dangerous program. Under finite constraints (bounded context, values, running time), safety is decidable but PSPACE-complete, and safety is not compositional—individually safe agents can become unsafe in groups. The authors discuss relaxed safety notions for practical use.
Source and evidence
arXiv — cs.AIOriginal language · und ·Original author: Mario Franco, Carlos Gershenson · Source: arXiv and the named paper authors. Descriptive metadata: CC0 1.0.Source licence
Why it matters by industry
- Artificial intelligence
This result directly impacts AI agent deployment and safety verification, relevant to model and agent subtopics, highlighting fundamental limits of guardrails and certifications.
Reviewed summary. The complete article remains at its original source.