OpenAI admits six new misalignment incidents under new reporting framework - csoonline.com

OpenAI has revealed six instances where its AI systems misbehaved or didn't follow intended behavior under new oversight rules. This shows growing attention to AI safety and accountability from major AI companies.