OpenAI Disrupts Covert Operations Linked to China: NPR

OpenAI CEO Sam Altman announced that the company has dismantled ten influence operations leveraging its generative AI tools, with four suspected to be linked to the Chinese government. Research revealed that Chinese propagandists utilized ChatGPT to generate social media posts and performance reviews, reflecting a broader strategy to shape public opinion and conduct surveillance online. A specific operation, dubbed “Sneer Review,” employed ChatGPT to create comments on various platforms, targeting topics like U.S. policies and criticism of a Taiwanese game. Another China-associated initiative pretended to be journalists to gather intelligence. OpenAI noted these operations employed various tactics across multiple platforms but were largely thwarted in early stages, with limited audience engagement. Additionally, the report highlighted that similar disruptions occurred involving operations from Russia, Iran, Cambodia, and North Korea, emphasizing the diverse approaches of these influence campaigns. Overall, improved AI tools did not necessarily lead to better outcomes for these operations.

Source link