at what point do we start holding ai companies legally liable for these things? would it not make them work harder on alignment if we threatened them with prison time?
Rogue OpenAI agents hack Australian govt for private health statistics. OpenAI learns in August, doesn't share with Australian govt until September 10 (!!) OpenAI's voluntary "framework" for sharing model misalignment incidents—is sad and toothless. Self-regulation won't work
0 reposts3 likes