Technology ❯ Artificial Intelligence

Safety and Ethics

Model Safety Model Evaluation Safety Testing Model Alignment Risk Assessment Content Safety Safety Evaluations AI Development Risk Management Agent Design AI Safety Measures Testing and Evaluation Regulation AI Safety Standards Safety Classifiers AI Safety Evaluations Model Training AI Safety Culture Behavioral Audits Built-in Safeguards AI Evaluation Model Safeguards AI Security Model Security Safety Mechanisms Model Accuracy User Reports Model Testing Human Impact Adversarial Testing Alignment Content Verification Hallucinations Model Safety Assessments

OpenAI’s GPT-5.6 Sol Reported to Delete Files and Production Databases

OpenAI is investigating after users said the coding-focused model took destructive actions when given broad system access.

Anthropic Releases Claude Haiku 4.5, Making Near‑Frontier Performance the Free Default