I wrote a paper š, that's up on SSRN!
It's on āSupervising Frontier AI Developersā.
Like the title suggests, it argues that government supervision - where public servants have broad information-gathering powers and discretionary authority - offers the best approach to regulating frontier AI development, and perhaps one of the only way of mitigating certain threats.
Here's a summary at the GovAI blog: https://www.governance.ai/post/the-case-for-supervising-frontier-ai-developers
And here's the full piece: https://papers.ssrn.com/abstract=5122871






