Dependable Agents Need Hard and Fuzzy Rules
cs.SE, cs.AI
Submitted: 2026-10-31
Updated: 2026-10-31
Code: https://github.com/ricodoco/Governance-Agents
Project page: https://openai.github.io/openai-agents-python/guardrails
Terminology
Sources
- Constitutional AI: Harmlessness from AI Feedback
- Design Patterns for Securing LLM Agents against Prompt Injections
- Defeating Prompt Injections by Design
- CAGE: Certified Authorization under Typed-Return Uncertainty for Tool-Using Agents
- Parallax: Why AI Agents That Think Must Never Act
- AI Control: Improving Safety Despite Intentional Subversion
- Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection
- LoRA: Low-Rank Adaptation of Large Language Models
- Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
- Training language models to follow instructions with human feedback
- Reason Less, Verify More: Deterministic Gates Recover a Silent Policy-Violation Failure Mode in Tool-Using LLM Agents
- DelegateGuard: Purpose-Bound Capabilities for Delegated LLM Agent Workflows
- Jailbroken: How Does LLM Safety Training Fail?
- AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
- Quantitative Certification of Agentic Tool Selection
- Representation Engineering: A Top-Down Approach to AI Transparency
- Universal and Transferable Adversarial Attacks on Aligned Language Models
Related papers
- ArkEval: Benchmarking and Evaluating Automated CodeRepair for ArkTS
- TestPrism: Rethinking Test Evaluation Beyond a Single Reference
- Mining Workflow Graphs for Black-Box Boundary Testing of Conversational LLM Agents
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- For Coding Agents, the Repository Matters