ChatGPT produced graphic violent images that shocked researchers

AI assistants like ChatGPT are supposed to have appropriate guardrails to stop people creating harmful content. However, they don't always work.

Raccontata damalwarebytes.com