AI Labs Need Guardrails, Not Empty Promises
Artificial intelligence companies face two related tests: keeping powerful systems safe and speaking honestly about product releases. I believe they are failing when voluntary promises replace oversight, or when “available
