Before such initiatives become enforceable laws, enterprises need to shape their AI strategies with the safety of AI agents ...