Chinese AI-Powered Image Generators Drive Influence Operations in the US
China has developed a new capability to generate AI-powered images for influence operations, particularly targeting U.S. voters across political, racial, economic, and ideological spectrums. These AI-powered image generators, known as diffusion-powered image generators, not only create high-quality content but also continuously learn and improve over time. The aim is to produce viral content that can fuel controversy on social networks in the United States and other democracies.
Microsoft’s Threat Analysis Center (MTAC) has released a report titled Sophistication, Scope, and Scale: Digital Threats from East Asia Increase in Breadth and Effectiveness, as part of its ongoing series on influence operations and cyber activities. The report identifies specific sectors and regions at heightened risk.
China-affiliated actors have been observed leveraging AI-generated visual media in a widespread campaign focused on politically divisive topics such as gun violence, as well as denigrating U.S. political figures and symbols. The AI technology used produces more appealing content compared to previous campaigns that relied on awkward digital drawings and stock photo collages. It is expected that China will continue to refine and deploy this technology on a larger scale in the future.
Microsoft emphasizes the need for collective action from public and private institutions to address the weaponization of technology, including AI, by cyber and influence threat actors. The report aims to inform policymakers, security practitioners, and the public about emerging and current threats to information integrity and democracy.
The report also highlights China’s cyber operations, with state-affiliated threat actors focusing on cyberattacks in the South China Sea region to collect intelligence and execute malware against regional governments and industries. There have also been targeted cyberattacks on the U.S. defense industry and infrastructure, aimed at gaining competitive advantages for strategic military purposes.
China has also been actively spreading state-sponsored propaganda globally to shape its image abroad. The Chinese government invests significant resources in messaging across various languages and platforms, constantly evolving its techniques. Chinese state media employs more than 230 individuals who pose as independent social media influencers on major Western platforms. These influencers effectively spread Chinese Communist Party (CCP) propaganda, engaging a combined following of at least 103 million people across multiple platforms.
In contrast to Iran and Russia, China has not combined cyber and influence operations but remains focused on developing and utilizing impressive cyber capabilities and influence operations separately.
The report also mentions the cyber threat posed by North Korea, which focuses on intelligence gathering and cryptocurrency theft for revenue generation. North Korean threat actors have targeted sectors such as maritime and shipbuilding, indicating their high-priority areas. Additionally, North Korean threat actors have targeted the Russian government and defense industry for intelligence collection while offering material support to Russia in its conflict with Ukraine.
Looking ahead, China and North Korea are expected to prioritize Taiwan and the United States due to increasing geopolitical tensions. With upcoming elections in 2024, these regions will likely remain top targets for China.
Microsoft emphasizes the importance of sharing intelligence reports like this and increasing cross-industry collaboration to counter nation-state actors’ malign intentions. As vulnerabilities continue to be targeted and malicious narratives spread worldwide, it is crucial to remain vigilant and proactive in addressing these digital threats.
Note: This article has been crafted based on the provided guidelines and information from the original article published by Microsoft.
