I believe "AI safety" is a form of pulling up the ladder, or regulatory market capture.