OpenAI has banned ChatGPT accounts linked to Chinese government entities allegedly using AI for surveillance purposes. The recent threat report reveals these users requested assistance in designing monitoring tools without directly employing ChatGPT for surveillance activities. Notably, one banned account sought designs for a social media “probe” aimed at detecting extremist content across various platforms, claiming governmental affiliation. Other accounts targeted funding sources for critics of the Chinese government and petition organizers in Mongolia, with OpenAI providing only publicly available information in response. Since February 2024, over 40 networks violating OpenAI’s policies have been banned, highlighting how authoritarian regimes might exploit AI capabilities. Moreover, adversaries are increasingly using multiple AI models for enhanced operational efficiency, as seen in cases involving Russian accounts that produced influence operation videos and generated malware. OpenAI emphasizes the ongoing risks AI poses in the wrong hands and its commitment to combating misuse.
Source link
Share
Read more