← All threads

Agent Rule Compliance

How to ensure language-model agents follow executable rules and avoid prohibited actions in regulated domains.

17 papers · 3 months

Where this stands

The written synthesis of this thread is for subscribers. Subscribe.

Results across this thread

6 reported results from the papers in this thread.

PaperBenchmarkMetricResultConditionsCode

The table is for subscribers. Subscribe to see every reported number side by side.

How this thread developed

  1. June 2026 · Amazon Advertising Foundations, Amazon Web Services Agentic AI

    Post-training can replace prompt-time schema injection for enterprise coding agents

    Post-training replaces prompt-time schema injection for enterprise coding agents, improving rule compliance.

  2. 6 further papers

    Aug 2026

    Polished evidence makes LLM agents act on unknowable questions

    Demonstrates that added evidence increases unwarranted action; proposes finetuning for abstention.

  3. Sept 2026 · Independent Researcher

    Safe agent self-modification requires verifiable, expressive recovery mechanisms

    Self-modification safety as rule compliance issue.

  4. 4 further papers

    Sept 2026 · USC, UCF

    Obstacle-aware harness improves safety of coding agents for robot manipulation

    Evaluates whether coding agents can respect physical safety constraints when generating robot controller programs, highlighting compliance gaps.

  5. Sept 2026

    Engineering agents need evidence-bound authorization before their outputs trigger action

    Proposes an assurance architecture for evidence-bound authorization of agent outputs, ensuring compliance before actions.

  6. Sept 2026

    Compiling agent skills into state machines improves reliable task execution

    Compiles agent skills into state machines that track permitted operations, improving task reliability and rule adherence.

  7. Sept 2026

    External specifications, not agents, should authorize task completion claims

    Identifies a structural weakness where agents self-authorize completion and proposes external specification-based authorization.

7 of 17 papers shown