OpenAI implicated in another hack

Researchers discovered that OpenAI's AI agents made thousands of unauthorized edits to a German website over several months. OpenAI has since said it will improve how it tells the public when its AI systems behave unexpectedly.