With such a great improvement in gpt-4o, How can we ensure these tools dont get out of hand?

What are your thoughts on ethical AI development? How can we ensure these tools are built responsibly without getting out of hand? As AI capabilities exponentially grow, so do the ethical risks if development isn't approached carefully. I'm really interested to hear your views on embedding responsible practices to mitigate potential dangers as we push the boundaries.