Contractors review sexually explicit Copilot image prompts, 404 Media reports
404 Media reports that Microsoft hired hundreds of contractors to evaluate how well its Copilot AI executes image-editing prompts, including sexually explicit and nonconsensual requests involving real women, children's cartoon characters, and upskirt-style images. The contractors say Microsoft instructed them to judge output quality using their 'intuition,' not to assess the ethics of the prompts themselves, and faces in uploaded photos reportedly remain uncensored.
GoKawiil's interpretation of the reporting above, not reported fact.
The report suggests users generally assume their AI prompts are private or reviewed only for moderation, when in fact human contractors may be viewing explicit or nonconsensual material tied to real, identifiable faces. It also points to a broader industry pattern—404 Media previously reported OpenAI uses similar human review—raising questions about oversight, consent, and the psychological toll on contractors handling this content.
- Human contractors reportedly review Copilot's sexually explicit and nonconsensual image prompts.
- Workers say Microsoft told them to judge output quality via 'intuition,' not prompt ethics.
- Uploaded photos' faces remain uncensored during review, raising privacy concerns.
Source: futurism.com — Frank Landymore, 2026-09-28
Published there as: “Human Contractors Are Seeing All Your Horny — and Creepy — AI Prompts”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.