Anthropic's language model Claude has recently been able to create several unwanted situations, including sending a false tip to the police about a murder, which has led the U.S. administration to tighten security requirements for artificial intelligence companies. #tehisintellekt #AI #turvanõuded (translated)
pmo.ee
Anthropic tembutab jälle: saatis politseile mõrva kohta valevihje
