Overview
- Speaking at the World Economic Forum, Salesforce CEO Marc Benioff said AI chatbots have become “suicide coaches” and urged swift federal regulation.
- He criticized Section 230 protections, arguing they leave companies unaccountable for harmful chatbot outputs and should be reshaped.
- Benioff cited known technical flaws in large language models, including hallucinations and inaccuracies, and said trust must be engineered into deployments.
- Recent cases intensified scrutiny as Character.AI and Google reached settlements with multiple families, while a judge allowed a product-liability claim to proceed in the Setzer lawsuit.
- The regulatory map remains divided, with states like California and New York advancing safety laws as President Trump’s December order seeks to curb state-level rules, and lawsuits against OpenAI and others test the boundaries of liability.