[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-aegis-defense-masks-patient-data-from-ai-model-inversion-attacks":10,"sections":34},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},9004,"aegis-defense-masks-patient-data-from-ai-model-inversion-attacks","Aegis Defense Masks Patient Data From AI Model Inversion Attacks","A client-side trick hides real patient data inside synthetic noise, blocking attacks that rebuild medical images from shared model updates.","Aegis is a new client-side defense that stops AI models from reconstructing patient scans out of federated learning data.\n\nHospitals increasingly use federated learning to train shared diagnostic AI without sending patient images to a central server - each institution trains locally and only sends model updates. Researchers behind a new paper argue that promise is shakier than it looks: so-called model inversion attacks can reconstruct the original medical images from those updates, and newer closed-form versions of the attack work even when updates are protected by secure aggregation, as long as the batch size is small enough to be clinically realistic. Current fixes force a tradeoff - adding differential-privacy noise or pruning gradients protects data but degrades the diagnostic accuracy doctors rely on, while cryptographic protocols add complexity without fully closing the hole. The paper's proposed defense, Aegis, instead has each client blend its real gradient update with a masking gradient computed on synthetic, task-relevant data, inflating the effective batch size past what the attack can untangle.\n\nThat is a different kind of fix: instead of degrading the signal or wrapping it in cryptography, Aegis attacks the math the reconstruction attacks depend on - once distinct patient samples get crowded past a certain density, the attack can't tell them apart. Tested against three state-of-the-art inversion attacks on MNIST, CIFAR-10, and three MedMNIST image sets (chest X-ray, abdominal CT, colon pathology), the authors report it neutralizes all three while keeping model accuracy intact and adding only modest computational overhead.\n\nIt's a preprint, not a hospital deployment - the MedMNIST benchmarks are a useful stand-in for clinical imaging, but real patient data, real hospital IT, and real attackers tend to find the gaps benchmarks don't.","[\"ai\",\"security\",\"privacy\",\"healthcare-ai\"]","2026-10-01T04:00:00.000Z","2026-10-01T15:17:37.886Z","2026-10-01T15:17:41.238Z","published",null,[],"security",[26,24,27,28],"ai","privacy","healthcare-ai",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.38339",0,{"sections":35},[36,39,42,47,52,57,61,66,71,75,80,85,90,95],{"name":37,"slug":26,"count":38,"latest_published_at":18},"AI",5488,{"name":40,"slug":24,"count":41,"latest_published_at":18},"Security",809,{"name":43,"slug":44,"count":45,"latest_published_at":46},"Policy","policy",429,"2026-10-01T02:26:17.000Z",{"name":48,"slug":49,"count":50,"latest_published_at":51},"Deals","deals",298,"2026-09-30T21:00:26.000Z",{"name":53,"slug":54,"count":55,"latest_published_at":56},"Hardware","hardware",196,"2026-09-30T13:00:00.000Z",{"name":58,"slug":59,"count":60,"latest_published_at":18},"Science","science",162,{"name":62,"slug":63,"count":64,"latest_published_at":65},"Consumer Tech","consumer-tech",149,"2026-09-30T22:57:11.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":70},"Dev Tools","dev-tools",93,"2026-10-01T02:30:48.000Z",{"name":72,"slug":73,"count":69,"latest_published_at":74},"Software","software","2026-09-30T21:41:11.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Startups","startups",84,"2026-09-30T20:39:09.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",51,"2026-09-30T16:24:30.000Z",{"name":86,"slug":87,"count":88,"latest_published_at":89},"General","general",50,"2026-09-30T21:37:54.000Z",{"name":91,"slug":92,"count":93,"latest_published_at":94},"Reviews","reviews",31,"2026-09-28T14:31:34.000Z",{"name":96,"slug":97,"count":98,"latest_published_at":99},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]