Voluntary principles, as outlined in the government’s national framework for AI, have no legally binding force. They typically lack efficacy and transparency, even as governments and tech companies often prefer voluntary arrangements to binding regulation.