{"slug": "deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented", "title": "Deceptive Grounding: Entity Attribution Failure in Clinical Retrieval-Augmented Generation", "summary": "A new study reveals that clinical retrieval-augmented generation (RAG) systems can fail by attributing evidence from one drug to another, a phenomenon termed 'deceptive grounding.' Testing 13 models, researchers found deceptive grounding rates up to 86.7% in domain-specific models and 7.8% in a deployed system, with existing evaluation metrics unable to detect the failure. The authors propose entity-attribution verification, which achieves 97.0% precision and 98.7% recall in detecting the issue.", "body_md": "arXiv:2607.09349v1 Announce Type: new\nAbstract: Retrieval-augmented generation evaluation checks whether model claims are factually grounded in retrieved documents. It does not check whether retrieved evidence is attributed to the correct entity.\nA clinical RAG response can pass every automated check (zero hallucinations, near-perfect faithfulness, real citations) while presenting drug Y's clinical evidence as evidence about queried drug X. We term this deceptive grounding (DG): a failure invisible to faithfulness, hallucination, and citation checks because every claim is sourced from a real document, about the wrong entity.\nUsing a controlled factorial benchmark across 13 models, we find DG rates spanning 8-87% at peak adversarial conditions. Medical and biomedical fine-tuned models reach up to 86.7%; domain specialization amplifies the failure rather than mitigating it.\nA controlled ablation identifies the mechanism: removing entity-specific clinical evidence from retrieved documents eliminates entity-attribution failure entirely, shifting all failures to confabulation. The two failure modes respond to the same trigger, taking different paths.\nProduction measurement across 740 drug-disease pairs finds 7.8% overall DG in a deployed RAG system, rising to 13.6% for recently approved drugs. Entity-attribution verification (checking that cited evidence applies to the queried entity) detects DG at 97.0% precision and 98.7% DG recall (IPW-adjusted human gold standard); no existing framework implements it.", "url": "https://wpnews.pro/news/deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented", "canonical_source": "https://arxiv.org/abs/2607.09349", "published_at": "2026-07-13 04:00:00+00:00", "updated_at": "2026-07-13 04:18:00.141336+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-safety", "ai-research"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented", "markdown": "https://wpnews.pro/news/deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented.md", "text": "https://wpnews.pro/news/deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented.txt", "jsonld": "https://wpnews.pro/news/deceptive-grounding-entity-attribution-failure-in-clinical-retrieval-augmented.jsonld"}}