The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Amazon has begun training its generative AI models on Twitch streamers' content—including videos, chats, and images—by default, unless users manually opt out. This policy change has sparked significant backlash over privacy and intellectual property concerns, as creators' content is used without explicit consent, raising ethical and rights issues.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves an AI system (Amazon's generative AI models) trained on Twitch users' content without explicit opt-in consent, which is a direct use of user-generated content for AI training. This use without explicit permission constitutes a violation of intellectual property rights and possibly privacy rights, fulfilling the criteria for harm under (c) violations of human rights or breach of obligations under applicable law protecting intellectual property rights. The harm is realized as users have reacted negatively and the default setting implies prior use of their content. Hence, this is an AI Incident rather than a hazard or complementary information.[AI generated]