Free Press Writer Admits Using ChatGPT to Edit Viral Photo From El‑Sayed Event
The admission prompted calls for clearer newsroom disclosure rules and intensified scrutiny of AI image‑detection methods.
Overview
- A post by Free Press writer Olivia Reingold that listed designer items seen at an election‑night gathering went viral after she attached an edited photo that drew about 3.2 million views.
- Vanity Fair and other observers flagged visual oddities in the image and reported that Google DeepMind’s SynthID labeled the file as made or edited with OpenAI tools.
- On Thursday Reingold acknowledged she ran the photo through ChatGPT to “brighten and sharpen” it and published what she said were her original, unedited images while saying the observation about designer items still stood.
- Analysts pointed out concrete changes between the viral image and Reingold’s originals—including different bag quilting and a crisper clasp plus altered shoe details—and several journalists called the failure to disclose the AI edit a serious ethical lapse while The Free Press had not issued a public policy response.
- The episode highlights how AI upscaling can hallucinate detail, creates new verification challenges for newsrooms, and carries heightened reputational stakes given the political sensitivity around Sen.‑elect Abdul El‑Sayed and prior scrutiny of the outlet.