I hear about an "AI safety" split, but what exactly does that mean? I feel like there's the existential risk "AI is going to escape and take over the world" fol

I hear about an "AI safety" split, but what exactly does that mean? I feel like there's the existential risk "AI is going to escape and take over the world" folks and the more mundane "AI is going to spout racism and propaganda" types. Is that the division that I'm hearing about?