Overview of the Discovery
WIRED has identified around 400 applications and websites, predominantly associated with Chinese-language developers, that ostensibly function as regular AI tools. However, these platforms harbor concealed features that allow users to generate violent and sexually abusive AI-generated images and videos, including deepfakes.
Key Dimensions of the Issue
Reputational Risk
- Impact on AI Companies: The revelation of these hidden features poses a severe reputational risk to AI companies. Allegations of facilitating modern slavery through such applications can significantly damage the credibility and public image of these enterprises.
Technological Products
- AI Applications and Websites: The core of this issue lies in the technological products themselves—AI applications and websites that are being misused to create harmful content.
Geographical Influence
- China's Role: The involvement of Chinese-language developers underscores China's significant influence on global AI policies. This could have far-reaching implications for international regulatory frameworks.
Ethical and Legal Non-compliance
- Deepfakes and Sexual Abuse: The primary threat is the creation and dissemination of violent and sexually abusive content. This not only raises ethical concerns but also exposes users and developers to potential legal actions.