OpenAI has uncovered an increasing number of Chinese-linked groups using its artificial intelligence technology for covert operations and influence campaigns on social media.
In a report released Thursday, the San Francisco-based company emphasized that while the activities were relatively small in scale and reach, the tactics were becoming more sophisticated.
According to the report, several ChatGPT accounts were suspended after being found generating politically and geopolitically charged content. Examples include:
- Criticism of a Taiwan-themed video game
- False accusations against a Pakistani activist
- Posts targeting U.S. foreign aid and referencing tariffs imposed by Donald Trump
In other cases, China-linked actors used AI for open-source intelligence gathering, script editing, and developing tools for cyberattacks and social media automation.
Another influence operation aimed to inject polarizing content into the U.S. political discourse using AI-generated texts and images.
OpenAI, recently valued at $300 billion following a $40 billion funding round, regularly publishes transparency reports on the misuse of its technology, including the generation of misleading content or the creation of harmful tools.
