It's Easy To Thwart A.I. Safety.

The article presents a report from the Canegie Mellon University in Pittsburgh, Pennsylvania and the Center for A.I. Safety in San Francisco, Calfornia which showed how to circumvent artificial intelligence (AI) safety systems using any chatbot or open source system to generate harmful information. Research findings indicate concern over potential spread of false information or disinformation, disagreements among AI companies, and danger of illegal online selling and buying of firearms.