There is no regulatory scenario where OpenAI doesn't get whatever they want. They are more a part of the US administration than not at this point. You would have to ignore all evidence to suggest that behaving responsibly has any effect on political outcomes in 2026.