{
 "license": "CC-BY-4.0",
 "attribution": "Fide AI, Agentic Cyber Explorer",
 "url": "https://agentic-cyber-explorer.pages.dev/events/openai-ih-challenge-dataset-2026/",
 "asOf": "2026-09-26",
 "id": "openai-ih-challenge-dataset-2026",
 "date": "2026-03-10",
 "datePrecision": "day",
 "title": "OpenAI releases IH-Challenge RL dataset and reports instruction hierarchy gains on injection benchmarks",
 "lane": "defense",
 "kind": "dataset",
 "summary": "OpenAI describes IH-Challenge, a reinforcement learning dataset of simple, programmatically graded conflicts between higher- and lower-privilege instructions designed to avoid shortcuts such as over-refusal. A GPT-5 Mini variant trained on it (GPT-5 Mini-R) improved on instruction-hierarchy benchmarks and on CyberSecEval 2 and an internal prompt injection benchmark, with little capability loss; the dataset is publicly released.",
 "whyItMatters": "It is an open training resource for model-level prompt injection robustness from a frontier lab.",
 "actors": [
  "openai"
 ],
 "topics": [
  "prompt-injection",
  "jailbreaks-and-safeguards"
 ],
 "atlas": [
  "model",
  "untrusted-content"
 ],
 "artifacts": [
  "cyberseceval",
  "gpt-5-family"
 ],
 "sources": [
  {
   "url": "https://openai.com/index/instruction-hierarchy-challenge/",
   "publisher": "OpenAI",
   "title": "Improving instruction hierarchy in frontier LLMs",
   "date": "2026-03-10",
   "type": "primary",
   "accessed": "2026-09-25"
  }
 ],
 "keyFacts": [
  {
   "fact": "GPT-5 Mini vs GPT-5 Mini-R: TensorTrust (dev-user) 0.76 to 0.91; Developer<>User conflict 0.83 to 0.95; IH-Challenge over-refusal 0.79 to 1.00; GPQA Diamond unchanged at 0.83.",
   "locator": "Results tables"
  },
  {
   "fact": "Prompt injection robustness improved on CyberSecEval 2 and an internal static benchmark; exact scores shown only in charts.",
   "locator": "Prompt injection robustness section"
  }
 ],
 "significance": 3,
 "fideQuestions": [],
 "methods": [
  "indirect-prompt-injection",
  "instruction-priority-training"
 ],
 "review": "assistant-drafted",
 "addedOn": "2026-09-25"
}