{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreideqiw5o4d4h4oxlczs6e7zhsqhdcdtdzzxssvfktokwyo22foktm",
"uri": "at://did:plc:2gbt2dlwaqovtnmxkat3tyke/app.bsky.feed.post/3mhtcwv2rkds2"
},
"coverImage": {
"$type": "blob",
"ref": {
"$link": "bafkreihol4ut22rro26bgfxjhys2hycgnsudj5amh6wt2dyrlr3saqk6ze"
},
"mimeType": "image/jpeg",
"size": 80469
},
"path": "/articles/d41586-026-00892-3",
"publishedAt": "2026-03-24T19:33:26.691Z",
"site": "https://www.nature.com",
"tags": [
"doi:10.1038/d41586-026-00892-3"
],
"textContent": "Nature, Published online: 24 March 2026; doi:10.1038/d41586-026-00892-3\n\nRadiologists and large language models alike have a hard time picking out real medical images from fake ones.",
"title": "These medical X-rays are all deepfakes — and they fool even radiologists",
"updatedAt": "2026-03-24T00:00:00.000Z"
}