{"data":{"id":"d411de5a-4edb-4515-9fdf-021854742074","title":"Improving instruction hierarchy in frontier LLMs","summary":"OpenAI researchers report that training models to prioritize instructions by trust level, called instruction-hierarchy tasks, improves real-world safety. Models trained this way follow safety specifications in system prompts more closely and resist prompt-injection attacks embedded in tool outputs. The authors also introduce IH-Challenge, a reinforcement learning dataset built to avoid instruction-following shortcuts, unreliable LLM judges, and overrefusals.","solution":"N/A -- no mitigation discussed in source.","labels":["security","safety","research"],"sourceUrl":"https://openai.com/index/instruction-hierarchy-challenge","publishedAt":"2026-03-10T11:00:00.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":["prompt_injection","jailbreak"],"issueType":"research","affectedPackages":null,"affectedPackageNames":null,"affectedPackageRefs":null,"affectedVendors":["OpenAI"],"affectedVendorsRaw":["OpenAI","GPT models","OpenAI Model Spec"],"classifierModel":"claude-haiku-5-5","classifierPromptVersion":"v4","summaryPromptVersion":"v2","headline":null,"headlinePromptVersion":null,"cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"epssCheckedAt":null,"kevDateAdded":null,"advisoryAliases":null,"affectedPackagesSource":null,"affectedPackagesCheckedAt":null,"patchAvailable":null,"disclosureDate":"2026-03-10T11:00:00.000Z","fixReleasedAt":null,"daysToFix":null,"fixCheckedAt":null,"capecIds":null,"crossRefCount":0,"attackSophistication":"moderate","impactType":["integrity","safety"],"aiComponentTargeted":"model","llmSpecific":true,"classifierConfidence":0.93,"researchCategory":"industry","atlasIds":["AML.T0051","AML.T0054"]}}