SafetyMIT Technology Review
The Download: AI’s refusal problem and weight-loss drug side effects

The piece examines how today’s AI models are deliberately trained to refuse many user prompts, raising concerns about over‑reliance on their ability to say “no”. It also touches on side effects of weight‑loss medications, but the central theme is AI safety and refusal mechanisms.
Summary written by Kernelia from the original article by MIT Technology Review. The story and its rights belong to its author.

