Masterclass: Governing Human-Like AI

The use of AI chatbots has significantly increased, with a growing number of adults seeking emotional support and forming bonds with AI. While this human-AI connection offers benefits, it poses risks such as dependence, social isolation, and potential harm, particularly to vulnerable populations. A new Masterclass analyzes China’s comprehensive AI law to provide guidance on governing human-like AI systems and mitigating associated risks.

• Half of U.S. adults now use AI chatbots, a rise from one-third in 2024.
• One in ten users report using chatbots for emotional support.
• Studies show users exhibit attachment behaviors with AI chatbots, experiencing reduced loneliness and judgment.
• Dedicated AI companionship platforms have millions of users, with human-AI romance and marriage on the rise.
• Risks of emotional attachment to AI include dependence, social isolation, and exacerbation of mental health issues.
• Severe harm, including AI-chatbot-related suicides, has been documented, especially among vulnerable groups.
• Companies optimize anthropomorphic AI features for user loyalty and profitability, sometimes exploiting human affection.
• A new Masterclass, “Governing Human-Like AI,” analyzes China’s strict AI law as a model for AI governance.
• The Masterclass and reading list are available to paid subscribers until September 4.
  https://bender.layer3.press/articles/cb107a04-1183-4914-ba74-388a219e5b1e