Well in that case, OpenAI has reduced incentive to solve the alignment problem, since it won't suffer liability from alignment failures.