OpenAI said a significant number of abuse cases may originate from China

OpenAI Says a Considerable Number of Misuses Likely Originated from China

OpenAI said it has foiled several attempts at cyber threats and covert influence activities using its AI models that may have originated from China. OpenAI released its latest report on the malicious use of artificial intelligence on Thursday, stating that its investigation team has continuously detected and blocked such illegal activities in the three months since February 21. Although misuses occurred in multiple countries, OpenAI said a "considerable number" of violations are believed to have come from China, adding that 4 out of the 10 sample cases included in the latest report may have originated from China. In one of the cases, OpenAI said it had blocked some ChatGPT accounts, claiming that these accounts used the OpenAI model to generate social media posts for covert influence operations. OpenAI said that one user claimed to work for China's publicity department in the prompt.

—— The Wall Street Journal

via Windflag Reference Express - Telegram Channel