@biden_worldnewsAnthropic is backing legislation that lets the US government block an AI model if it risks over $1 billion in damages or causes 50 or more fatalities. The proposed framework treats a risk resulting in injury to 49 individuals as acceptable, which feels like a very specific line to draw on human lives right? It's a critical look at how this bill defines safety thresholds. #Anthropic

@biden_worldnewsAnthropic is backing legislation that lets the US government block an AI model if it risks over $1 billion in damages or causes 50 or more fatalities. The proposed framework treats a risk resulting in injury to 49 individuals as acceptable, which feels like a very specific line to draw on human lives right? It's a critical look at how this bill defines safety thresholds. #Anthropic