{
  "id": 3284,
  "url": "https://arxiv.org/abs/2606.02276v1",
  "title": "Cross-modal linkage risk in clinical vision-language models",
  "summary": "Vision-language models (VLMs) trained on paired chest radiographs and radiology reports learn a shared embedding space that can preserve instance-level image-report correspondence. This poses a privacy risk in settings where radiographs and reports are deliberately kept separate after acquisition, such as image-only data sharing or access-controlled reports, because a de-identified image may be re-linked to its original narrative report through cosine similarity alone. We formalized this as imag",
  "authors": "Soroosh Tayebi Arasteh, Mahshad Lotfinia, Sven Nebelung, Daniel Truhn",
  "category": "research",
  "topics": "privacy-surveillance,healthcare",
  "orgs": null,
  "regions": null,
  "published_at": "2026-06-01T14:01:46.000Z",
  "fetched_at": "2026-07-14T16:30:09.959Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3284",
  "original_url": "https://arxiv.org/abs/2606.02276v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}