The Midas Project hosts several websites documenting AI companies’ behavior; I think the most relevant one is Seoul Tracker which tracks how well AI companies are living up to their commitments at the Seoul summit.
SaferAI gives AI companies risk management ratings.
The first two of these are solo projects by Zach Stein-Perlman.
There are a few orgs doing things like this:
AI Lab Watch rates AI companies on their safety procedures along various dimensions.
AI Safety Claims Analysis critically reviews AI companies’ safety claims.
The Midas Project hosts several websites documenting AI companies’ behavior; I think the most relevant one is Seoul Tracker which tracks how well AI companies are living up to their commitments at the Seoul summit.
SaferAI gives AI companies risk management ratings.
The first two of these are solo projects by Zach Stein-Perlman.