OpenAIbào gào报告：yǒu rén有人yòng用ChatGPTzuò做wǎng luò网络mǒ hēi抹黑，měi guó美国yì yuán议员dān xīn担心bèi被yòng lái用来dǎ yā打压yì yì rén shì异议人士
OpenAI recently released a report stating that they discovered a Chinese law enforcement officer using ChatGPT as a 'work tool' to write plans, edit texts, and attempt some online influence operations, such as spreading unfavorable content about Japanese politician Sanae Takaichi on social media and pretending to be foreign residents filing complaints.
ChatGPT rejected some of the requests, and OpenAI later suspended the user's access.
The report also mentioned that this user impersonated a US immigration officer to intimidate Chinese dissidents in the US, requested AI to generate fake court documents to get others' accounts banned on platforms, and even created fake obituaries and tombstone photos claiming a dissident had died; these contents later appeared online.
Some US lawmakers commented that using AI for such purposes would make cross-border suppression faster and more like an 'assembly line'.
OpenAI stated that the report was published to warn everyone to prevent the malicious use of AI.