Skip to main content
See every side of every news story
Published • loading... • Updated

Microsoft Updates Copilot AI Image Generation Policy With Human Reviewers

Contractors say they are being asked to judge sexual and illegal image prompts as Microsoft uses human review to measure Copilot output quality.

  • On Monday, 404 Media reported that human contractors hired to improve Microsoft Copilot are reviewing user prompts and uploaded photos, seeing uncensored images including sexually explicit material.
  • According to documents obtained by 404 Media, Microsoft instructed contractors to rely on their intuition when evaluating if the AI followed prompts rather than flagging offensive content.
  • Contractors describe being inundated with disturbing requests, including upskirt photos and pro-anorexia content. One contractor said they "recoiled" after reviewing such material on internal message boards.
  • A Microsoft spokesperson told 404 Media the company uses customer data to improve products and enforce its code of conduct. Contractors, often based in the Global South, are not paid to moderate exploitative content.
  • This practice is widespread; 404 Media recently reported that OpenAI also employs hundreds of contractors to read ChatGPT conversations. Jacob Andreou, Microsoft's executive vice president for Copilot, told Business Insider the company focuses on AI "diffusion" risks.
Insights by Ground AI

13 Articles

Think freely.Subscribe and get full access to Ground NewsSubscriptions start at $9.99/yearSubscribe

Bias Distribution

  • 83% of the sources lean Left
83% Left

Factuality Info Icon

To view factuality data please Upgrade to Premium

Ownership

Info Icon

To view ownership data please Upgrade to Vantage

404 Media broke the news in United States on Monday, September 28, 2026.
Too Big Arrow Icon
Sources are mostly out of (0)

Similar News Topics

News
Feed Dots Icon
For You
Search Icon
Search
Blindspot LogoBlindspotLocal