The Chinese government is reported to have used local artificial intelligence firms to conduct sophisticated propaganda campaigns, according to researchers at Vanderbilt University who reviewed documents from one such company, GoLaxy.
Nextgov/FCW reported that the researchers found that GoLaxy created data profiles on over 2,000 American political and thought leaders and that there was evidence that the AI firm monitored thousands of right-wing influencers and journalists.
National Security Priorities
Founded in 2010 as an affiliate of the Chinese Academy of Sciences, GoLaxy allegedly executed campaigns against Hong Kong and Taiwan and appears to follow Beijing’s national security priorities.
The researchers noted that the documents suggest the firm has collaborated with senior intelligence, party and military officials in China, although no direct government control has been confirmed.
Addressing the Triple D of Bad Actor Threats – Disproportionality, Disinformation and Deception is among the most anticipated panel discussions at the Potomac Officers Club’s upcoming 2025 Intel Summit on Oct. 2. Reserve your slot today.
Gray Zone Conflict
Brett Goldstein, former head of the Defense Digital Service and a member of the Vanderbilt research team, described the operation as a new “level of gray zone conflict” that warrants closer scrutiny. He made the remarks during DEF CON in Las Vegas alongside retired Gen. Paul Nakasone, who leads Vanderbilt’s National Security Institute.
Nakasone, former National Security Agency director and seven-time Wash100 recipient, emphasized that AI has enabled influence operations “at a speed and a scale we’ve never seen before.”
Influence in AI Sector
These revelations about GoLaxy’s activities emerge as China’s influence in the AI sector expands. In July, a Digital Science report found that China is leading global AI research, holding over 40 percent of global citation attention and dominating AI-related patents. The report also noted that Chinese research is becoming increasingly influential in the U.S., the United Kingdom and the European Union.
Additionally, a Reuters report last year revealed that researchers from three Chinese institutions, including two linked to the People’s Liberation Army, used Meta’s open-source LLaMA model to develop a military chatbot called ChatBIT. The team claimed that their system performs nearly as well as OpenAI’s GPT-4 for military dialogue and question-answering tasks, although its deployment status remains unclear.