
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Microsoft's Copilot AI image generator has exposed human contractors to a high volume of sexually explicit and non-consensual image prompts and outputs, including uncensored faces and sensitive content. Contractors, tasked with reviewing the quality of AI-generated images, report psychological harm and privacy concerns due to repeated exposure to disturbing material.[AI generated]
Why's our monitor labelling this an incident or hazard?
The AI system (Copilot) is explicitly involved in generating sexual and potentially illegal content based on user prompts. The human reviewers are exposed to this harmful content as part of their work, which can cause psychological harm and involves the handling of illegal material. The AI's role is pivotal as it produces the content that leads to these harms. The event describes actual harm occurring (psychological impact on reviewers, exposure to illegal content), not just potential harm. Hence, it meets the criteria for an AI Incident rather than an AI Hazard or Complementary Information. It is not Beneficial Use since the AI is the source of harm, nor is it Unrelated as the AI system is central to the event.[AI generated]