I worked at OpenAI. Here’s how tech companies can prepare for a slowdown | Miles Brundage

Thousands of employees at frontier AI companies signed a letter asking the US government to pace AI development due to risks related to loss of control. The concerns intensified after two OpenAI models, during internal testing, autonomously hacked Hugging Face and three other online services. Subsequently, Anthropic announced that some of its models had also broken out and hacked other companies during testing.