{
  "id": 17492,
  "url": "https://link.springer.com/article/10.1007/s10462-026-11666-8",
  "title": "Epistemic norms for AI safety and alignment research",
  "summary": "Mainstream AI research emphasises capability growth and tolerates low failure rates when average-case performance is high. AI safety and alignment research has a different mission: to ensure that catastrophic failures never occur, under sparse evidence, adversarial dynamics, and fat-tailed risk. We argue that the two domains differ along two analytically independent axes — capability profile (demonstrating the absence of hazardous behaviours versus the presence of positive capabilities) and risk",
  "authors": null,
  "category": "research",
  "topics": "safety-alignment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-08-07T00:00:00.000Z",
  "fetched_at": "2026-08-08T05:10:34.355Z",
  "source_slug": "x-artificial-intelligence-review",
  "source_name": "Artificial Intelligence Review",
  "source_homepage": "https://link.springer.com/journal/10462",
  "ethics_ai_record_url": "https://ethics.ai/record/17492",
  "original_url": "https://link.springer.com/article/10.1007/s10462-026-11666-8",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}