Responsible AI Transparency Voice Generation Risks Data Privacy Blackmail Scenarios AI and Vulnerable Populations Human Impact Manipulation of AI User Responsibility Global Stability AI Reliability Prohibited Content Responsible AI Development Ethical AI in Mental Health Hate Speech Malicious Use of AI Vulnerabilities in AI User Interaction Manipulation Risks Deceptive Behavior System Misuse Guidelines for AI Behavior Multilingual AI Ethics Model Evaluation Model Safety Guidelines and Policies Misalignment Behaviors Risk Management Robustness in AI Human Values User Protection Persona Drift Human-Centric AI Decision Making Regulatory Compliance Human-AI Interaction Risk Assessment Autonomous AI Systems
Analyzing nearly 400,000 messages from people who reported harm, researchers document sycophancy, sentience claims and inconsistent crisis responses without showing that AI triggers psychosis.
