AI leaders say they want safety, but on whose terms?

Leaders from OpenAI, Anthropic, and Google have expressed willingness to discuss AI safety, though with certain conditions.